assessment of the performance of light-emitting diode roadway lighting technology
TRANSCRIPT
Assessment of the Performance of Light-Emitting Diode Roadway Lighting Technology
http://www.virginiadot.org/vtrc/main/online_reports/pdf/16-r6.pdf RONALD B. GIBBONS, Ph.D. Center Director Center for Infrastructure-Based Safety Systems YINGFENG “ERIC” LI, Ph.D., P.E. Research Associate Center for Infrastructure-Based Safety Systems JASON E. MEYER Research Associate Center for Advanced Automotive Research Virginia Tech Transportation Institute
Final Report VTRC 16-R6
Standard Title Page—Report on State Project
Report No.:
VTRC 16-R6
Report Date:
October 2015
No. Pages:
74
Type Report:
Final Contract
Project No.:
RC00043
Period Covered:
March 2012- September 2015
Contract No.:
Title:
Assessment of the Performance of Light-Emitting Diode Roadway Lighting
Technology
Key Words:
Light-Emitting Diode; LED
Roadway Lighting
Author(s):
Ronald B. Gibbons, Ph.D., Yingfeng “Eric” Li, Ph.D., P.E., and Jason E. Meyer
Performing Organization Name and Address:
Virginia Tech Transportation Institute
3500 Transportation Research Plaza (0536)
Blacksburg, VA 24061
Sponsoring Agencies’ Name and Address:
Virginia Department of Transportation
1401 E. Broad Street
Richmond, VA 23219
Supplementary Notes:
Abstract:
This study, championed by the Virginia Department of Transportation (VDOT) Traffic Engineering
Division, involved a thorough investigation of light-emitting diode (LED) roadway lighting technology by
testing six types of roadway luminaires (including housing and all components enclosed) in a laboratory
environment and on the field over a 2-year period.
The results showed that LED luminaires exhibited superior lighting and related qualities compared
to high-pressure sodium luminaires. Different photometric characteristics were found among LED
luminaires of different designs, indicating a careful selection considering light distribution and illuminance
level is necessary for individual lighting applications. During the first 2 years of operation, the average
light loss for the LED luminaires was 6% based on laboratory testing. The study also found that
implementing LED technology systematically will result in a return on investment between 3.25 and 5.76
for different scenarios over a 25-year period due to savings in maintenance and energy consumption.
The study resulted in the VDOT LED Roadway Luminaire Specification document and developed
recommendations relevant to VDOT’s implementation of LED technology.
FINAL REPORT
ASSESSMENT OF THE PERFORMANCE OF LIGHT-EMITTING DIODE ROADWAY
LIGHTING TECHNOLOGY
Ronald B. Gibbons, Ph.D.
Center Director
Center for Infrastructure-Based Safety Systems
Virginia Tech Transportation Institute
Yingfeng “Eric” Li, Ph.D., P.E.
Research Associate
Center for Infrastructure-Based Safety Systems
Virginia Tech Transportation Institute
Jason E. Meyer
Research Associate
Center for Advanced Automotive Research
Virginia Tech Transportation Institute
Project Manager
Benjamin H. Cottrell, Jr., P.E., Virginia Transportation Research Council
Virginia Transportation Research Council
(A partnership of the Virginia Department of Transportation
and the University of Virginia since 1948)
Charlottesville, Virginia
October 2015
VTRC 16-R6
ii
DISCLAIMER
The project that is the subject of this report was done under contract for the Virginia
Department of Transportation, Virginia Transportation Research Council. The contents of this
report reflect the views of the author(s), who is responsible for the facts and the accuracy of the
data presented herein. The contents do not necessarily reflect the official views or policies of the
Virginia Department of Transportation, the Commonwealth Transportation Board, or the Federal
Highway Administration. This report does not constitute a standard, specification, or regulation.
Any inclusion of manufacturer names, trade names, or trademarks is for identification purposes
only and is not to be considered an endorsement.
Each contract report is peer reviewed and accepted for publication by staff of the
Transportation Research Council with expertise in related technical areas. Final editing and
proofreading of the report are performed by the contractor.
Copyright 2015 by the Commonwealth of Virginia.
All rights reserved.
iii
ABSTRACT
This study, championed by the Virginia Department of Transportation (VDOT) Traffic
Engineering Division, involved a thorough investigation of light-emitting diode (LED) roadway
lighting technology by testing six types of roadway luminaires (including housing and all
components enclosed) in a laboratory environment and on the field over a 2-year period.
The results showed that LED luminaires exhibited superior lighting and related qualities
compared to high-pressure sodium luminaires. Different photometric characteristics were found
among LED luminaires of different designs, indicating a careful selection considering light
distribution and illuminance level is necessary for individual lighting applications. During the
first 2 years of operation, the average light loss for the LED luminaires was 6% based on
laboratory testing. The study also found that implementing LED technology systematically will
result in a return on investment between 3.25 and 5.76 for different scenarios over a 25-year
period due to savings in maintenance and energy consumption.
The study resulted in the VDOT LED Roadway Luminaire Specification document and
developed recommendations relevant to VDOT’s implementation of LED technology.
1
FINAL REPORT
ASSESSMENT OF THE PERFORMANCE OF LIGHT-EMITTING DIODE ROADWAY
LIGHTING TECHNOLOGY
Ronald B. Gibbons, Ph.D.
Center Director
Center for Infrastructure-Based Safety Systems
Virginia Tech Transportation Institute
Yingfeng “Eric” Li, Ph.D., P.E.
Research Associate
Center for Infrastructure-Based Safety Systems
Virginia Tech Transportation Institute
Jason E. Meyer
Research Associate
Center for Advanced Automotive Research
Virginia Tech Transportation Institute
INTRODUCTION
Roadway and street lighting across the nation consumes a large amount of energy and is
responsible for significant tonnage of carbon dioxide (CO2). The DOE estimated that the 26.5
million streetlights in the US consumed as much electricity each year as 1.9 million households,
and generated greenhouse gas emissions equivalent to that produced by 2.6 million cars.1
Currently, the majority of the lamps used on American roads are traditional gas-discharge lamps,
such as high-pressure sodium (HPS), low-pressure sodium (LPS), mercury vapor (MV), and
metal halide (MH). In addition to their high power consumption, the service lives of such lamps
are greatly affected by an array of factors.2
As the nation continues to make efforts in conserving energy and reducing CO2 emission,
and as state and local transportation agencies strive to save operational costs, more efficient
roadway lighting technologies are becoming increasingly appealing. Among the various
emerging technologies, solid-state lighting (SSL) differs from other lighting technologies in that
it is based on light-emitting diodes (LEDs) or organic LEDs (OLEDs) instead of filaments,
plasma, or gases.3 Among various technologies, LED is currently gaining popularity for general
illumination applications as many challenges still remain in the adaptation of other SSL
technologies. LEDs typically have a much longer service life and provide higher luminaire
efficacy than traditional sources. Their light output is also a much broader spectrum than other
sources, meaning that the light appears to be white and provides proper color appearance. This
feature can result in an improved visual performance for the same design light level of traditional
lighting technologies.
2
Over the past decade, LED lighting has experienced significant growth in roadway
lighting applications. However, it is still in its early stages due to reasons related to immaturity
of the technology and unfamiliarity with the technology at state and local transportation
agencies. Estimates in 2010 suggested that only approximately 3% of the streetlight lamps
across the nation were LED. In contrast, HPS technology represents 80% of the streetlight lamps
currently used in the United States.4
While providing the benefits of energy efficiency and superior light quality, however,
LEDs have changed the lighting industry. Traditionally, the lighting industry has been
dominated by a very few companies who had extensive experience in providing products that
would be able to withstand the exterior environment. Now, small, typically electronics
companies have started the development of lighting products. These systems may have issues
with the durability of the luminaires and the ease of the installation of the luminaires. Similarly,
the luminaires must match the requirements of the application to provide a proper light
distribution while reducing glare and uplight. With the advent of all of these less experienced
companies and even the production of the new technology by traditional lighting companies, it is
important that careful evaluations of the luminaires be made before the full scale implementation
of the technology is undertaken.
To develop a comprehensive understanding of this fast evolving technology, stakeholders
have launched various studies across the nation. The U.S. Department of Energy (DOE), for
example, is performing a large-scale field assessment of SSL performance for general
illumination in exterior and interior applications, known as the SSL Technology Demonstration
GATEWAY program.5, 6
Some state DOTs and municipalities also conducted similar studies to
understand LED technology.7, 8, 9, 10
Past experience with LED technology in roadway lighting
has suggested the following:
The light efficacy of LED roadway lighting systems is comparable to that of HPS in
roadway lighting applications, but is improving rapidly and is expected to exceed that
of other traditional technologies significantly in the near future. Many studies,
particularly those conducted in the early 2000s, suggested that LED lighting systems
had a comparable efficacy compared to HPS technology on a lumen-per-watt basis.
However, some recent studies reported significant increases in efficacy for LED
systems. DOE expects that a 200 lm/W efficacy for LED will be achieved in the
future - doubling to tripling that of conventional incandescent lighting.11
Replacing traditional luminaires with LED luminaires of comparable wattages is not
likely to result in significant cost savings. However, most studies showed that LED
systems with much lower wattages were capable of emitting lights meeting minimum
design standards, therefore yielding more significant cost savings.12
Most studies concluded that the light quality of LED systems, such as light color,
distribution, perception, and ground illuminance, was superior to that of traditional
lighting technologies.6 User surveys also suggested that most roadway users
preferred LED lighting to traditional lighting.
3
There is a wide range of LED manufacturers and models commercially available for
roadway lighting. Some LED lighting products vary considerably in cost and lighting
performance (e.g., color rendering, efficacy, life span, and light distribution).
Many previous LED lighting studies were based on relatively short-term lighting data
measurements. As such, few studies have resulted in a thorough understanding of
LED lumen maintenance over time based on field measurements.
Recognizing the critical needs for energy conservation and better lighting, the Virginia
Department of Transportation (VDOT) Traffic Engineering Division (TED) championed this
study to comprehensively assess LED roadway lighting technology and LED performance over
time. The results of this research provide valuable insights in key performance differences
between standard HPS and LED technologies as well as among LED luminaires of different
designs. The findings also fill in a knowledge gap regarding how LED luminaires perform
differently in a laboratory environment and in the field. The knowledge developed from this
research served as the basis for the development of the VDOT LED Roadway Luminaire
Specification document and associated implementation recommendations.
PURPOSE AND SCOPE
The primary objectives of this study were as follows:
Develop a comprehensive understanding of LED lighting performance for roadway
lighting applications based on laboratory and field evaluations.
Identify performance improvements and cost savings associated with a potential
adoption of LED technology for roadway lighting and related purposes.
Develop a specification document and recommendations relevant to the adoption of
LED systems for roadway lighting at VDOT.
During the study, the research team tested and monitored five different LED luminaire
designs for 2 years and compared their performance characteristics to those of standard HPS
luminaires.
METHODS
Overview
Five major tasks were performed to meet the study objectives:
1. Conduct a literature review to summarize previous findings relevant to the
performance of LED technology in roadway lighting applications in comparison with
traditional roadway lighting systems.
4
2. Conduct rigorous laboratory evaluations of LED systems to determine their lighting
performance metrics.
3. Conduct multi-year field evaluations to determine the lumen maintenance and field
lighting performance of LED lighting technology over time.
4. Perform economic analysis to determine potential energy and cost savings associated
with LED lighting systems as compared with existing VDOT roadway lighting
systems.
5. Develop recommendations and specifications for using LED lighting systems at
VDOT-maintained facilities.
A laboratory test was conducted first and focused on key luminaire performance metrics
such as power consumption, light output, and spectral performance in a controlled environment.
The luminaires were then tested at a VDOT park-and-ride facility for field assessment in an
effort to understand LED lighting performance over time. For comparison, the Virginia Tech
Transportation Institute (VTTI) team also performed laboratory and field evaluations of VDOT’s
standard HPS fixtures. During data analysis, all outdoor measurements of the LED systems were
corrected based on ambient horizontal and vertical illuminance levels. In addition, all LED
illuminance measurements were normalized to a standard temperature (i.e., 25°C) assuming that
a reduction of each degree Celsius in ambient temperature coincides to a light output increase by
0.25%.
LED System Selection
Based on previous research experience and VDOT recommendations, the research team
contacted a list of reputable LED lighting vendors to acquire sample systems for evaluation. At
the end of the process, interested vendors provided sample LED systems of different designs.
For benchmarking purposes, the research team also used three 250 W HPS luminaires of the
same design from a single manufacturer. Table 1 lists the LED systems evaluated during this
study followed by photographs of the luminaires in Figure 1. Note that Design B had a
manufacturer-related Correlated Color Temperature (CCT) of 5000+300, which was much
higher than that of other LED luminaires and was in the cool white range.
Table 1. Luminaires Evaluated During the Study
Design Mfg.
Year
Mfr.
Rated
Watt
Correlated
Color
Temp.
Mfr. Rated
Lumen
Weight
(lb) Qty. LED Design Feature
HPS 03/2012 250 - - - 3 N/A
Design A 04/2012 195 4300 4452 - 6 Exposed LED optic array
Design B 2012 120 5000+300 8985 25 6 Three-panel folding design with large
LED sources
Design C 05/2012 148 4000
- 45 6 Three large LED sources with
conventional layout
Design D 2012 150 4000 9285-13890 25 6 Exposed, elongated LED optic array
Design E 08/2011 200 4000 - 32 6 Exposed LED optic array
5
Figure 1. LED and HPS Systems Used in Study
Laboratory Evaluation
VTTI conducted two rounds of laboratory evaluations as part of this study, both
following a similar process:
1. Initial laboratory testing. After obtaining the luminaires, the VTTI team performed
the initial laboratory evaluation of the luminaires to compare the different LED
luminaire designs with each other and with HPS. The initial testing entailed
mounting two sample luminaires (labeled No. 1 and No. 2) of each manufacturer in a
laboratory facility for an initial “burning-in” time of approximately 100 hours and
then installing the luminaires individually in an outdoor VTTI test facility for detailed
performance assessment.
2. Second laboratory testing. After the luminaires were tested in the VDOT test bed for
2 years, the research team retrieved the luminaires and conducted final laboratory
testing for comparison with the results of the initial laboratory testing. During the
second round of testing, the research team first collected data from each luminaire in
the same condition as when it was retrieved from the field installation (i.e., dirty
condition) and then cleaned the luminaires and tested them again. The Design C (1)
HPS 250
Design E
Design A Design D
Design B
Design C
6
luminaire was not properly retrieved after the field testing and therefore that
luminaire was not included in some of the comparison analyses. In addition, due to
site conditions, only HPS (2) was evaluated on the test bed and in the second round of
laboratory testing.
The VTTI outdoor testing facility consisted of a light pole fitted with an adjustable
bracket that could accommodate most luminaire types. The research team defined a
measurement grid extending 6 m behind, 13 m in front of, and 20 m to each side of the luminaire
(Figure 2). Luminaires were mounted at a height of 30 ft (9.1 m) because a majority of
conventional roadway luminaires at VDOT are installed at a height between 30 and 45 ft (9.1
and 13.7 m).
Figure 2. Overhead View of Laboratory Evaluation Grid (m) and Measurement Method
During laboratory testing, the research team collected the following measurements:
Horizontal illuminance, measured with a Minolta T-10 illuminance meter on the
pavement, facing up, at the center of each cell in the 20 x 40 m grid, as shown in
green in Figure 3.
Vertical illuminance, measured using a Minolta T-10 illuminance meter affixed to a
mobile cart, mounted 1.5 m from ground level, as shown in red in Figure 3. During
the data collection, the meter was aimed along the roadway in the direction of the
luminaire; in the left half of the grid, the illuminance meter was aimed parallel to the
grid facing the right, and in the right half of the grid, the meter was aimed parallel to
the grid facing the left.
Light trespass, measured as vertical illuminance along the front and back edges of the
grid with the meter mounted at 1.5 meters from the ground level and facing the
luminaire side, as shown in violet in Figure 3.
7
Electrical power usage, measured with a Yokogowa WT 110 power meter. The
research team waited at least 15 min after the LED luminaire was powered on before
taking this measurement to avoid the potential effects of in-rush current. For each
HPS measurement, the research team waited at least 30 min after the HPS luminaire
stabilized.
Spectral power distribution (SPD), which was measured using an Ocean Optics
S4000 spectroradiometer with a Teflon integrating sphere acceptance optic. The
research team measured only the relative irradiance as the research team’s interest
was the relative power concentration by wavelength. Irradiance is defined as the
amount of radiant flux hitting or passing through a unit area of a surface. Relative
irradiance measures the shape of the light spectrum but not the absolute magnitude,
which allows a user to determine whether there is more light at one wavelength than
another. To facilitate comparisons, the SPD results for different luminaires were
normalized to the same scale.
The initial laboratory testing was conducted at night, between 9 P.M. and 1 A.M. in May
2012, and the second laboratory testing was conducted between February and March 2015, also
at night.
Figure 3. Horizontal and Vertical Illuminance Measurement Systems
Collecting horizontal and vertical illuminance measurements over the 20 x 40 m grid
required 800 readings for each luminaire, which was time-consuming. To improve efficiency
and accuracy, the research team developed an automated data acquisition application in the
National Instruments® LabVIEW software environment. During the data collection, the
automated application collected continuous illuminance readings from the Minolta T-10
illuminance meter for 2 s at each location and then wrote the mean illuminance into a comma-
separated values (CSV) file. To enable real-time data validation, the application interface
included measurement visualization windows as well as buttons that allowed values to be redone
or deleted upon faulty measurements (Figure 4).
8
Figure 4. Illuminance Data Acquisition Application
Field Evaluation
Test Site Settings
During the field evaluation, all sample luminaires, including both LED and HPS systems,
were installed in a park-and-ride facility that VDOT designated as a lighting test bed between
September 2012 and September 2014. The test bed is in Woodbridge, Virginia, on the north side
of Telegraph Road, approximately 600 ft (183 m) northeast of Caton Hill Road or 2,000 ft (610
m) northwest of I-95. The facility is in the middle of a large wooded area with minimum
interference of environmental lighting from adjacent roadways and commercial and residential
developments.
Before the official opening of the parking facility in September 2012, 26 of the 33
luminaires acquired (five LED systems of each type and one HPS system limited by parking lot
lighting needs) were installed at a standard height of approximately 35 ft (10.6 m) for testing.
The performance of six of the installed luminaires, one of each type and all lab tested, was
monitored for 24 months. Figure 5 shows the locations of the luminaires installed in the test
facility. The circled luminaires are those for which manual measurements were collected. The
luminaires were operating every night from dusk to dawn, equivalent to a total operational period
of approximately 8,800 hours. Among the six tested luminaires, one Design D and one Design E
luminaires were installed directly above a bus station and subjected to different dirt/dust
condition than other luminaires in the parking lot.
9
Figure 5. Telegraph Road Evaluation Area Luminaires
Field Data Collection
During the field evaluation period, from September 2012 to September 2014, the research
team conducted field measurements at 3-month intervals, for a total of nine rounds of data
collection. During each visit, the research team collected measurements both manually and with
the automated data collection system. Before each data collection, the research team took
multiple ambient horizontal and vertical illuminance measurements for controlling ambient
factors during data analysis. The hourly temperatures during each data collection were later
obtained for the nearest weather station from the National Oceanic and Atmospheric
Administration (NOAA).
During each site visit, the research team manually collected the following data:
Horizontal illuminance, with the same equipment as used for laboratory data
collection, but on a smaller version of the laboratory grid. The field measurement
grid was 12 m x 36 m, with slight variations made to accommodate site conditions
such as curbs and sidewalks. For efficiency, the research team took measurements at
3-m increments.
Vertical illuminance, with the same equipment as used for laboratory data collection
at a height of 1.5 meters above the ground. The vertical illuminance measurements
were taken using the same field grid as the horizontal illuminance measurements.
CCT, with a Minolta CL-500 Illuminance Spectrophotometer measured directly
beneath each luminaire.
Beginning with the second round of data collection, VDOT supplied the research team
with a bucket truck, allowing the team to visually inspect and record the physical condition of
the luminaires. Each inspection included a general visual inspection for dirt buildup and wildlife
intrusions, luminaire temperature recording (ballast and LED components), and an observation
of the luminaires’ installation condition.
10
Due to construction at the Telegraph Road Park & Ride, data were not collected for the
high-pressure sodium luminaire during the September 2013 visit.
During this study, the research team also made an attempt to collect lighting
measurement data at a finer spatial resolution using the VTTI Roadway Lighting Mobile
Measurement System (RLMMS). RLMMS synchronizes several key lighting measurement
devices including four illuminance meters designed for horizontal illuminance measurement.
However, due to the parking lot settings and luminaire installation locations, it was difficult to
collect measurements at the exact same locations for all luminaires. As such, the RLMMS data
analysis yielded bias that was relatively significant and therefore the results were not included.
LED Life Cycle Cost Analysis
Relevant Lighting Inventory Data at VDOT
Interviews with VDOT engineers suggested that VDOT does not currently maintain an
accurate inventory of luminaires in the state. They indicated that the vast majority (> 95%) of
the luminaires used at VDOT are HPS systems. Figure 6 summarizes the information obtained
from VDOT staff during interviews relevant to the roadway luminaire composition at VDOT.
Table 2 and Table 3 further list VDOT-estimated lighting inventory and relevant maintenance
costs based on the most recent VDOT’s highway lighting needs assessment. The VDOT
Methods for Calculating Maintenance and Operations Needs: Highway Lighting – Asset 380
document13
details the procedures and assumptions used to obtain these estimates. Note that the
values in Table 2 and Table 3 are provided separately by VDOT officials and are more up-to-
date compared with those in the needs assessment document. Several pieces of critical
information needed for this cost analysis were derived based on these data.
Figure 6. VDOT-Maintained Luminaires by Wattage
VDOT Lighting Inventory
Sign Lighting (5%)Conventional Roadway
Lighting (80%)High-Mast Lighting (10%) Parking Lots (5%)
100W (5%) 150W (95%) 400W (50%) 1000W (50%)
250W (75%) 400W (25%)
Pedestrian: 150W (100%)
Parking Lots (60%)
150W (50%) 400W (50%)
Tunnel Lighting (2.5%) Roadway Lighting (95%)Under Bridge Lighting
(2.5%)
200W (10%) 310W (3%)150W (2%) 400W (35%)250W (50%)
150W (50%) 400W (50%)
11
Table 2. Estimates of VDOT Light Inventory (2014 Data)
Light Type VDOT District Interstate Primary Secondary Total
Conventional
light (each pole
is counted as 1)
Bristol 44 0 0 44
Salem 688 753 0 1,441
Lynchburg 0 324 0 324
Richmond 660 624 0 1,284
Hampton Roads 5,526 1,474 92 7,092
Fredericksburg 1,500 420 0 1,920
Culpeper 0 50 0 50
Staunton 0 45 0 45
Northern Virginia 10,044 4,937 2,890 17,871
Conventional light subtotal 18,462 8,627 2,982 30,071
High mast light
(each pole is
counted as 1)
Bristol 87 0 0 87
Salem 13 0 0 13
Lynchburg 0 0 0 0
Richmond 60 0 0 60
Hampton Roads 100 13 0 113
Fredericksburg 300 0 0 300
Culpeper 0 0 0 0
Staunton 13 0 0 13
Northern Virginia 345 205 144 694
High mast light subtotal 918 218 144 1,280
Sign light (each
luminaire is
counted as 1)
Bristol 36 0 0 36
Salem 27 5 10 42
Lynchburg 0 21 0 21
Richmond 881 479 0 1,360
Hampton Roads 1,007 196 0 1,203
Fredericksburg 695 90 10 795
Culpeper 0 27 0 27
Staunton 60 65 0 125
Northern Virginia 10,025 4,480 503 15,008
Sign light subtotal 12,731 5,363 523 18,617
Grand Total 32,111 14,208 3,649 49,968
Table 3. FY16 Lighting Maintenance & Operational Needs Lights $27,253,178
Ancillary Structure Maintenance - Lighting $5,826,494
Ancillary Structure Replacement - Conventional Lights $5,793,155
Ancillary Structure Replacement - High Mast Lights $2,950,556
Conventional Lighting Re-lamp/Elect Repair $1,324,206
High Mast Lighting Re-lamp/Elect Repair $41,444
Highway Lighting Power Bills $3,498,667
Sign Lighting Lifecycle Replacement $750,417
Sign Lighting Re-lamp/Elect Repair $668,535
Turnkey Asset Maintenance Services (TAMS)* $538,721
Underground Utilities Replacement $5,860,982
*TAMS refer to a specific type of contacts at VDOT performing routine, ordinary and
preventive maintenance of highway system and its assets.
12
Cost-Benefit Analysis Period and Scenarios
During this study, the research team used an analysis period of 25 years as suggested by
VDOT officials and according to manufacturer warranted LED luminaire operational life
required by VDOT. The research team examined a number of potential scenarios pertaining to
how VDOT would acquire and implement LED technology.
Currently, the two most common scenarios for State DOTs to obtain LED luminaires are
through leasing or purchasing:
Leasing. Many manufacturers offer LED luminaire leasing services. In general,
there can be two different options for leasing LED luminaires:
Provide LED luminaires at a lower or no cost to state DOTs but harvest a portion
of or entire energy cost savings for a certain period of time.
Provide LED luminaires at a lower or no cost to state DOTs but charges a
predetermined fee for each luminaire for a certain period of time.
Discussions with VDOT officials suggested that VDOT intends to fully acquire LED
luminaires instead of operating rented luminaires when starts concerting HPS
roadway lighting to LED. As such, this cost-benefit analysis did not consider the
leasing scenario.
Purchasing. This analysis only considered the scenario where VDOT purchases all
LED luminaires.
The researchers examined the following luminaire replacement scenarios:
Retrofitting. Many manufacturers offer LED luminaires that can be readily retrofitted
into existing luminaire housings. However, studies suggested that retrofitting existing
housings with LED luminaires did not necessarily result in more significant cost
savings over time.14
In addition, retrofitting luminaires does not enable full utilization
of the state-of-the-art LED roadway lighting technologies considering that LED
luminaires have very different optical control and thermal performance. Retrofitting
could also void manufacturers’ warrantees on the existing lighting systems and result
in liability issues.14
Furthermore, due to foreseeable technology improvements, it is
reasonably certain that the LED products on the market in the near future will be
considerably more efficient, physically different, and not mechanically compatible
with current luminaires. As such, the discussions with VDOT officials suggested that
VDOT would not be interested in the retrofitting scenario as well.
Replacing. This scenario assumes that VDOT will replace the existing fixtures with
new LED luminaires, using the existing poles/structural supports. It is the goal of
VDOT to ultimately convert all existing roadway luminaires to LED systems. The
most likely process for the conversion would be project by project. However, to
13
accurately account for this process in the cost analysis, it is necessary to obtain
information regarding future lighting projects on VDOT roadways. Such information
is extremely difficult to obtain due to data availability, reliability of planned project
schedules, and funding availability at VDOT. For simplicity, the research team
assumed the following replacement scenarios during the cost-benefit analysis:
Scenario 1 (S1): replacing all luminaires with LED luminaires at once. This
scenario can be considered as a baseline for comparison and better understanding
of the cost-benefits.
Scenario 2 (S2): phasing out within 5 years. This scenario assumes that
traditional luminaires will be phased out within a 5-year period and replaced with
an equal number of LED luminaires. Old luminaires approaching their designed
service lives will be replaced first.
Scenario 3 (S3): phasing out within 10 years. This scenario assumes that
traditional luminaires will be phased out within a 10-year period and replaced
with an equal number of LED luminaires. Old luminaires approaching their
designed service lives will be replaced first.
In summary, the research team considered the following scenarios during this cost-benefit
analysis:
Luminaire acquisition method: purchasing
Luminaire replacing method: replacing entire fixtures. The luminaires will be
replaced in the following three scenarios:
Replace all luminaires at year 1.
Replace all luminaires within a 5-year period.
Replace all luminaires within a 10-year period.
Cost-Benefit Analysis Factors and Formulae
The following is a list of the energy-related factors included in the cost-benefit analysis:
Current electricity cost ($/kWh) for DOT (CEC). Conversations with VDOT officials
suggested that it would be difficult to obtain an accurate estimate of a unit electricity
cost for roadway lighting at VDOT for a number of reasons. VDOT uses several
different billing mechanisms depending on the agreements with power companies,
location, and the service provider. Lighting fixtures installed recently tend to use
power meters while many older fixtures are covered by fixed flat-rate bills per service
(a number of luminaires within an agreed area). Based on the total estimated power
bills (Table 3) and roadway lighting wattage (see the following section) obtained
from VDOT, the research team estimated a power cost of $0.043 per kWh, which is
slightly lower than the rate at other state DOTs (e.g., 0.046 at MnDOT14
).
14
Future energy cost increase factor (ECI). The U.S. Energy Information
Administration (EIA) projects that the electricity price will continuously grow
following the previous growth trends.15
Currently, EIA projects that the U.S. retail
residential price for electricity will increase by 1.1% in 2015 and by 1.8% in 2016. In
addition, EIA data show that the residential electricity price increased annually by 3.6
% over the past decade on average (Figure 7). Based on the above information, the
research team estimated a 2% annual ECI taking inflation into consideration.
Figure 7. U.S. Residential Electricity Price
15
Rebates/incentives for LED lighting. To accelerate the use of energy-efficient
lighting technologies, many federal and state agencies offer incentives for
implementing LED roadway lighting.16
In addition, some utility companies also offer
rebates to customers for the use of LED luminaires.17
However, such incentive and
rebate programs usually change over time and will be typically phased out as LED
luminaires gain more popularity. Therefore, such incentives are difficult to quantify
for a long-term cost-benefit analysis and the research team did not include this factor
in the study.
The following is a list of the luminaire-related cost factors:
Number of existing luminaires (NEL) on VDOT roadways by type. Based on the
information obtained from VDOT, the research team estimated the number of
luminaires on VDOT roadways by wattage as listed in Table 4. To develop the
estimates, the research team assumed that 10% of the conventional light poles
included dual heads and an average of 6 luminaires are installed on each high-mast
light pole. Further, the research team assumed a 15% moderate wattage variation of
HPS luminaires according to the laboratory testing results and available studies.18, 19
15
Table 4. Estimated Luminaires by Wattage on VDOT Facilities (2014 Data)
Type/Wattage Conventional
Light
High-Mast
Light
Sign
Light Total
Total Watt
(Nominal)
Total Watt
(15% Variation)
100 W
931 931 93,100 107,065
150 W 2,343
17,686 20,029 3,004,350 3,455,003
200 W 3,227
3,227 645,400 742,210
250 W 17,088
17,088 4,272,000 4,912,800
310 W 968
968 300,080 345,092
400 W 12,460 3,840
16,300 6,520,000 7,498,000
1000 W
3,840
3,840 3,840,000 4,416,000
Total 36,086 7,680 18,617 62,383 - -
Total Watt (Nominal) 10,552,930 5,376,000 2,746,000 - 18,674,930 21,476,170
Total Watt (15% Variation) 12,135,870 6,182,400 3,157,900 - 21,476,170 -
Number of additional luminaires that VDOT plans to install during the analysis
period. Conversations with VDOT officials suggested that they have no plans in the
foreseeable future to add large numbers of lighting fixtures.
Current and future annual luminaire operating hours (LOH). VDOT operates
luminaires from dusk to dawn every day without light curfews. Based on the sunrise
and sunset time in 2014, the total annual lighting hours are estimated as 4,324 hours.
For simplicity, this study used 4,000 hours as the annual lighting hours.
Conversations with VDOT officials suggested that VDOT currently does not have
plans on increasing or reducing this operation time in the near future.
Current LED luminaire cost (LEDC). Based on the luminaire price obtained from the
vendors, the unit costs of the tested LED luminaires ranged from $1.8/W to $5.7/W
when ordering in large quantities (e.g., 100 luminaires). This analysis used the
average unit cost of the evaluated luminaires which was $3.45/W.
Current HPS lamp cost. This cost is included in the annual HPS re-lamping and
electric repair costs (see Table 3). Note that this analysis only considered existing
HPS fixtures assuming no new roadway lighting fixtures are planned based on
conversations with VDOT officials.
Future LED luminaire price reduction factor (LPRF). Currently, the LED
technology is still in its early maturing stage and it is expected that the price of LED
luminaires will continue dropping rapidly in the near future. A recent study showed
that the average LED luminaire price is expected to drop by about 30% between 2014
and 2017, with an average annual reduction factor of 10%.20
Knowing that the price
drop factor is not linear, the research team assumed a conservative LPRF:
LPRF = (9% - 9/25*t)
where t is the number of years into the 25-year analysis period from the base year
(i.e., 2015). With this factor, it is assumed that the average LED luminaire price in 10
years will be just over one-half of the current price, and the price in 25 years will be
16
about one-third of the current price. Note that DOE projected a more than 50% price
drop between 2013 and 2015, much higher than the value used for this analysis.11
Initial LED luminaire replacement wattage factor (LRWF). The research team did
not find third-party studies recommending an equivalent wattage for LED luminaires
when replacing HPS systems. Several LED luminaire manufacturers suggested a
replacement factor of 0.4 to 0.6 (e.g., a 100 W HPS luminaire can be replaced with a
40 to 60 W LED luminaire). Sample case studies showed examples where 250 W
HPS luminaires were replaced with LED luminaires ranging from 104 W (0.42) to
210 W (0.84) based on different design requirements and LED products.21
Users
should note that because different LED luminaires can have very different
photometric performances, it is not feasible to develop a universal wattage
replacement rate between HPS and LED luminaires. During this study, the wattage
of the LED luminaires used for comparison against the 250 W HPS systems ranged
from 120 W (0.48) to 200 W (0.8). Based on this information, the research team
assumed an LRWF of 0.6.
Luminaire operational life (LOL). The following are manufacturer-rated operational
lives for the studied luminaires:
HPS 250 W: the research team did not obtain a manufacturer-rated operational life
for the specific HPS luminaires evaluated, but the average service life for HPS
lamps was found to be 24,000 hours.22 Discussions with VDOT officials
suggested that VDOT re-lamps HPS luminaires every 1 to 2 years on normal
conditions and every 2 to 6 months for HPS luminaires installed on bridges
Design A LED luminaires: L70 (time required when the lumen output reaches
70% of the initial output) of 149,000 hours at 25°C
Design B LED luminaires: manufacturer calculated L70 of 914,000 hours and
reported L70 of 60,500 hours
Design C LED luminaires: L85 (time required when the lumen output reaches
85% of the initial output) of 50,000 hours
Design D LED luminaires: L70 of 100,000 hours at 25°C or 85,000 hours at 40°C
Design E LED luminaires: L70 of 80,000 hours at 25°C.
During this study, the research team used 100,000 hours for the LOL of LED
luminaires (LLOL) as this value is used in most state LED roadway luminaire
specifications and can be met by a majority of modern LED luminaires. The LOL for
HPS luminaires (HLOL) used in this study was 2 years based on VDOT feedback,
knowing that HPS luminaires on bridges are replaced much more frequently. In
addition, since most HPS lamps are replaced after complete lamp failures, salvage
values were not considered during this cost analysis.
17
LED luminaire efficacy increase factor (LEF). U.S. Department of Energy
projections suggested an increase in LED luminaire efficacy by 89% between 2013
and 2020.11
More conservative projections suggested efficacy increase factors
between 38% and 46% during the next decade from 2015 to 2025.19
Realizing this
factor is non-linear over time (i.e., the efficacy increase slows over time), the research
team used a simplified formula:
LEF = (6% - 6/25*t)
where t is the number of years into the analysis period. This LEF would result in a
50% efficacy increase in 10 years or a 100% increase in 25 years.
The following is a list of the installation- and maintenance-related cost factors:
Annual luminaire and ancillary structure maintenance cost. The annual maintenance
cost for HPS (HAMC; including costs of lamp replacement) is estimated as $0.96/W
(i.e., total lighting maintenance and operational costs with the exceptions of power
bills and underground utility replacement in Table 3 divided by total nominal wattage
in Table 4). The research team estimated an annual luminaire and ancillary structure
maintenance cost of $1.30/W for LED luminaires (LAMC) without luminaire costs
(i.e., total ancillary lighting structure maintenance and replacement costs in Table 3
divided by total nominal wattage in Table 4 and then by LRWF). Note that VDOT
currently does not conduct routine roadway luminaire cleaning and inspection. When
functioning properly, therefore, LED luminaires themselves (excluding ancillary
structures) are considered “maintenance free” since they do not require lamp
replacement.
Disposal cost per luminaire. Conversations with VDOT officials suggested that this
cost is typically included in the annual maintenance costs and cannot be separately
accounted for in a straightforward manner. As such, the research team did not
separate this cost from the maintenance costs.
Installation-related labor and traffic control costs. This study did not consider
additional costs associated with the labor and traffic control required for installing
LED luminaires. The researchers assumed that, regardless of luminaire types, new
lighting projects require the same installation costs. Replacing existing HPS
luminaires when they are functioning, however, results in additional installation-
related costs (particularly in the case of S1). However, conversations with VDOT
officials suggested that the installation related costs such as traffic control costs could
change significantly based on installation scenarios and whether it was done in-house.
Therefore, these costs were not considered in this study.
The following are factors and assumptions that are not included in the categories above:
Crash reduction savings and environmental benefits are not considered.
18
All costs are in 2015 dollars with an annual inflation rate of 2%.
Year is the basic time unit in this analysis for all scenarios and all LED luminaire
purchases are assumed at the beginning of each year. The timing of luminaire
replacement within individual years was not considered due to the complications
associated with potential changes of variables such as inflation, interest rate,
luminaire price, electricity price, and LED luminaire efficacy within a single year.
This study did not consider utilities-related costs including utilities maintenance,
relocation, and upgrades, with the assumption that the conversion to LED luminaires
would not result in significant increase in such costs.
Table 5 summarizes the factors used for this cost-benefit analysis.
Table 5. Summary of Cost Analysis Factors
Factor Variable Value
General Factors and Assumption
Analysis period (year) - 25
Analysis base year - 2015
Inflation rate r 2%
Energy-Related Costs
Current energy cost ($/kWh) CEC 0.043
Future energy cost increase factor ECI 2%
Total nominal wattage (W) TNW 18,674,930
Total HPS power consumption wattage (W) TCW 21,476,170
Current and future annual luminaire operating hours LOH 4,000
Luminaire-Related Factors
Current LED luminaire cost ($/W) LEDC 3.45
Future LED luminaire price reduction factor LPRF 9% - 9/25*t
Initial LED luminaire replacing wattage factor LRWF 0.6
LED luminaire operational life (hours) LLOL 100,000
HPS luminaire operational life (hours) HLOL 8,000
LED luminaire efficacy increase factor LEF 6% - 6/25*t
Luminaire and Ancillary Structure Maintenance Costs
Current annual maintenance cost for HPS luminaires ($/W/year) HAMC 0.96
Current annual maintenance cost for LED luminaires ($/W/year) LAMC 1.3
The research team used the following equations for the calculations:
Base scenario (without conversion to LED) annual energy cost for tth
year:
Eq. 1:
Base scenario annual maintenance cost for tth
year:
Eq. 2:
Costt,base,energy = 𝐶𝐸𝐶 × (1 + 𝐸𝐶𝐼)𝑡 × 𝑇𝐶𝑊 × 𝐿𝑂𝐻
Costt,base,maint = 𝐻𝐴𝑀𝐶 × 𝑇𝑁𝑊 × (1 + 𝑟)𝑡
19
Base scenario annual total lighting maintenance and operational cost for tth
year:
Eq. 3:
S1 (converting all HPS luminaires to LED in year 1) initial investment cost:
Eq. 4:
S1 annual energy cost after LED conversion for tth
year:
Eq. 5:
S1 annual maintenance cost after LED conversion for tth
year:
Eq. 6:
S1 annual total lighting maintenance and operational cost after LED conversion for tth
year:
Eq. 7:
S2 (converting all HPS luminaires in a 5-year period) annual LED luminaire cost for tth
year (t <=5):
Eq. 8:
S2 annual energy cost for tth
year:
Eq. 9:
S2 annual maintenance cost for tth
year:
Eq. 10:
S2 annual total lighting maintenance and operational cost for tth
year:
Eq. 11:
Costt,base,total = Costt,base,energy + Costt,base,maint
CostS1,initial = 𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × 𝐿𝐸𝐷𝐶
Costt,S1,energy = 𝐶𝐸𝐶 × (1 + 𝐸𝐶𝐼)𝑡 × 𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × 𝐿𝑂𝐻
Costt,S1,maint = 𝐿𝐴𝑀𝐶 × 𝑇𝑁𝑊 × (1 + 𝑟)𝑡
Costt,S1,total = Costt,S1,energy + Costt,S1,maint
Costt,S2,LED = 𝑇𝑁𝑊
5× 𝐿𝑅𝑊𝐹 × 𝐿𝐸𝐷𝐶 × (1 − 𝐿𝑃𝑅𝐹)𝑡
1
Costt,S2,energy = (𝑡
5× 𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × (1 − 𝐿𝐸𝐹))𝑡
1𝑡1 +
5−𝑡
5× 𝑇𝐶𝑊 ×
𝐶𝐸𝐶 × 1 + 𝐸𝐶𝐼 𝑡 if t <=5;
Costt,S2,energy = (𝑡
5𝑇𝑁𝑊/5 × 𝐿𝑅𝑊𝐹 × (1 − 𝐿𝐸𝐹)𝑡
1𝑡1 × 𝐶𝐸𝐶 × 1 + 𝐸𝐶𝐼 𝑡 if t
>5;
Costt,S2,maint = 𝑡
5× 𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × 𝐿𝐴𝑀𝐶 × (1 + 𝑟)𝑡 +
5−𝑡
5× 𝑇𝑁𝑊 × 𝐻𝐴𝑀𝐶 ×
(1 + 𝑟)𝑡 if t <=5;
Costt,S2,maint = 𝑇𝑁𝑊 × 𝐿𝐴𝑀𝐶 × (1 + 𝑟)𝑡 if t >5;
Costt,S2,total = Costt,S2,LED + Costt,S2,energy + Costt,S2,maint
20
S3 (converting all HPS luminaires in a 10-year period) annual LED luminaire cost for tth
year (t <=10):
Eq. 12:
S3 annual energy cost for tth
year:
Eq. 13:
S3 annual maintenance cost for tth
year:
Eq. 14:
S3 annual total lighting maintenance and operational cost for tth
year:
Eq. 15:
Development of VDOT Specification for Roadway LED Luminaires
As part of this study, the research team developed draft specifications of LED luminaires
for use on the VDOT-maintained roadways. The development was based on the LED evaluation
results of this study, a comprehensive understanding of existing standards and guidelines
relevant to the LED industry, interviews with VDOT officials, manufacturer and consultant
input, and a review of LED luminaire specifications of several sample state transportation
agencies.
In accordance with VDOT recommendations, the specification document was developed
as a special provision for quick implementation, with the intent of incorporation into the VDOT
Road and Bridge Specifications. The specification was intended to address requirements
relevant to the selection, testing, and installation of LED luminaires for use on VDOT facilities.
The document applies to conventional pole-mounted and wall-mounted luminaires with the
exception of high-mast luminaires and luminaires to be used for tunnel applications. The target
audience of the document is contractors performing VDOT lighting projects.
The developed specification (Virginia Department of Transportation Special Provision
for Light Emitting Diode [LED] Roadway Luminaires) was delivered to the Virginia
Transportation Research Council separately as a stand-alone product of this study.
Costt,S3,LED = 𝑇𝑁𝑊
10× 𝐿𝑅𝑊𝐹 × 𝐿𝐸𝐷𝐶 × (1 − 𝐿𝑃𝑅𝐹)𝑡
1
Costt,S3,energy = (𝑡
10× 𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × (1 − 𝐿𝐸𝐹))𝑡
1𝑡1 +
10−𝑡
10× 𝑇𝐶𝑊 ×
𝐶𝐸𝐶 × 1 + 𝐸𝐶𝐼 𝑡 if t <=10;
Costt,S3,energy = (𝑡
10𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × (1 − 𝐿𝐸𝐹)𝑡
1𝑡1 × 𝐶𝐸𝐶 × 1 + 𝐸𝐶𝐼 𝑡) if t
>10
Costt,S3,maint = 𝑡
10× 𝑇𝑁𝑊 × 𝐿𝑅𝑊𝐹 × 𝐿𝐴𝑀𝐶(1 + 𝑟)𝑡 +
10−𝑡
10× 𝑇𝑁𝑊 ×
𝐻𝐴𝑀𝐶 × (1 + 𝑟)𝑡 if t <=10;
Costt,S3,maint = 𝑇𝑁𝑊 × 𝐿𝐴𝑀𝐶 × (1 + 𝑟)𝑡 if t >10.
Costt,S3,total = Costt,S3,LED + Costt,S3,energy + Costt,S3,maint
21
RESULTS
Initial Laboratory Horizontal Illuminance
Horizontal illuminance is an indication of the distribution of light reaching the ground. It
also shows the level of uniformity of the luminaire’s output. Measurements were conducted for
two luminaires of each type, and the results for two luminaires of the same type were mostly
consistent. Readers should understand that higher absolute illuminance values may not
necessarily indicate good luminaire performance. The goodness of a luminaire’s photometric
performance is determined by multiple variables, such as horizontal/vertical illuminance,
uniformity, light distribution, CCT, and light loss over time.
Figure 8 through Figure 13 show the recorded horizontal illuminance values over the
laboratory measurement grid (40 x 20 m) for the six luminaire types evaluated, respectively. The
figures clearly suggest that the HPS system provided much higher horizontal illuminance values
at the focal center than the LED systems did. However, the horizontal illuminance relatively
concentrated beneath the luminaire and quickly decreased across the laboratory grid. Among the
LED systems, the Designs C and E systems exhibited the highest horizontal illuminance values,
with the Design E system showing more widespread light shed across the entire laboratory grid.
Figure 8. Laboratory Horizontal Illuminance – HPS 250W
22
Figure 9. Laboratory Horizontal Illuminance – Design A
Figure 10. Laboratory Horizontal Illuminance – Design B
Figure 11. Laboratory Horizontal Illuminance – Design C
23
Figure 12. Laboratory Horizontal Illuminance – Design D
Figure 13. Laboratory Horizontal Illuminance – Design E
Figure 14 shows the average horizontal illuminance over the laboratory grid for all
evaluated luminaire systems. Notice that although the Design A and D systems exhibited
relatively lower maximum horizontal illuminance, the horizontal illuminance values across the
entire measurement grid were much more uniform, resulting in average horizontal illuminance
levels similar to those of the Design C and E systems.
24
Figure 14. Average Laboratory Horizontal Illuminance
To better understand and compare the horizontal illuminance uniformity between the
LED and HPS systems and among the different LED designs, the research team selected the
illuminance readings along the 3-meter, 6-meter, and 9-meter grid lines in front of the luminaire,
as shown in Figure 15. If a roadway light were located directly above the edge line of the
rightmost lane, illuminance values along these three grid lines would roughly correspond to the
amount of light falling on the lane-marking lines for three adjacent traffic lanes.
Figure 15. Horizontal Illuminance at 3, 6, and 9 m Grid Lines
Figure 16 through Figure 18 compare the measured horizontal illuminance levels at the 3-
meter, 6-meter, and 9-meter laboratory grid lines, respectively. As the figures show, the
horizontal illuminance values of the HPS luminaire are much higher in the center of the 3-meter
grid but quickly approach the levels of other LED systems at approximately 9 m laterally away
from the luminaire. This is also shown in Figure 8 where much of the HPS light output
concentrates within a roughly 9-m circular area beneath the luminaire. When comparing across
the three grid lines, the horizontal illuminance values of the HPS system decrease significantly as
0
5
10
15
20
25
Illu
min
ance
(lu
x)
25
the distance increases transversely, with the peak values closely approaching those of the LED
systems.
Among the various LED designs, the Design C luminaire exhibited a relatively similar
horizontal illuminance distribution as that of the HPS systems. LED luminaires with LED optic
arrays in general had a much more spread-out light distribution.
Figure 16. Horizontal Illuminance at 3-Meter Grid Line
Figure 17. Horizontal Illuminance at 6-Meter Grid Line
0.0
20.0
40.0
60.0
80.0
100.0
120.0
19181716151413121110 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1011121314151617181920
Illu
min
ance
(lu
x)
Distance (Meter)
HPS 250 1_3
Design A 1_3
Design B 1_3
Design C 1_3
Design D 1_3
Design E 1_3
0.0
20.0
40.0
60.0
80.0
100.0
120.0
19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Illu
min
ance
(lu
x)
Distance (Meter)
HPS 250 1_6
Design A 1_6
Design B 1_6
Design C 1_6
Design D 1_6
Design E 1_6
26
Figure 18. Horizontal Illuminance at 9-Meter Grid Line
To understand the horizontal illuminance uniformity in a quantitative manner, the
research team calculated maximum and average uniformity ratios for the three grid lines based
on illuminance using a method similar to the concepts described in IES RP-8-14.23
In this
context, the ratios were calculated as:
Maximum Uniformity Ratio = Emax/Emin
and
Average Uniformity Ratio = Eavg/Emin
where Emax is the average of three continuous maximum horizontal illuminance readings; Emin is
the average of three continuous minimum horizontal illuminance readings; and Eavg is the
average of the horizontal illuminance readings along the entire grid line.
Figure 19 through Figure 21 illustrate the ratios for the three grid lines. Comparing the
figures, it is clear that the maximum uniformity ratio of the HPS system was much higher than
most LED systems at the 3-meter grid line. In addition, the Design C LED system (with three
large LED optics) exhibited a maximum uniformity ratio much higher than that of other LED
designs. In contrast, LED systems, such as Designs A, D, and E to a certain extent, showed a
better horizontal illuminance uniformity along all three grid lines and transversely between the
three grid lines.
Figure 19. Average and Max Uniformity Ratio at 3-Meter Grid Line
0.0
20.0
40.0
60.0
80.0
100.0
120.0
19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Illu
min
ance
(lu
x)
Distance (Meter)
HPS 250 1_9 Design A 1_9
Design B 1_9 Design C 1_9
Design D 1_9 Design E 1_9
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Average Uniformity Ratio Max Uniformity Ratio
HPS 250 (1) Design A (1)
Design B (1) Design C (1)
Design D (1) Design E (1)
27
Figure 20. Average and Max Uniformity Ratio at 6-Meter Grid Line
Figure 21. Average and Max Uniformity Ratio at 9-Meter Grid Line
Initial Laboratory Vertical Illuminance and Lighting Quality
Vertical Illuminance and Uniformity
Vertical illuminance is the amount of illuminance that lands on a vertical surface.
Vertical illuminance is an important roadway lighting metric as it is a reasonable criterion for
determining the amount of light landing on pedestrians. At 1.5 m from the ground, vertical
illuminance measurements also give an indication of the light that would adversely affect an
observer’s eyes creating glare. Similarly, the two luminaires of each design evaluated during this
study performed similarly, thus this section only discusses the results for one system for each
type.
Figure 22 through Figure 27 illustrate the vertical illuminance values of each different
luminaire system over the laboratory grid. As the figures suggest, the HPS system had much
higher peak values than the LED systems. However, the values decreased quickly across the
grid from the focal points. In general, the vertical illuminance values of the LED systems were
much more widespread on the grid, with the exception of the Design C Luminaires. These
results suggest that most LED systems outperform the HPS system in terms of reducing glare for
travelers.
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Average Uniformity Ratio Max Uniformity Ratio
HPS 250 (1) Design A (1)
Design B (1) Design C (1)
Design D (1) Design E (1)
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Average Uniformity Ratio Max Uniformity Ratio
HPS 250 (1) Design A (1)
Design B (1) Design C (1)
Design D (1) Design E (1)
28
Figure 22. Laboratory Vertical Illuminance – HPS 250 W
Figure 23. Laboratory Vertical Illuminance – LED Design A
Figure 24. Laboratory Vertical Illuminance – LED Design B
29
Figure 25. Laboratory Vertical Illuminance – LED Design C
Figure 26. Laboratory Vertical Illuminance – LED Design D
Figure 27. Laboratory Vertical Illuminance – LED Design E
30
Figure 28 shows the average vertical illuminance values of all evaluated luminaire
systems across the entire laboratory grid. The figure shows that the HPS systems generated
higher vertical illuminance on average, partly attributable to the high peak illuminance values.
Among the LED designs, Design E luminaires had the highest average vertical illuminance. The
Design A and D systems also emitted relatively high average vertical illuminance across the
evaluation grid, especially compared to their lower maximum vertical illuminance readings.
Figure 28. Average Laboratory Vertical Illuminance
The researchers also compared the vertical illuminance readings of the systems along the
three-, six-, and nine- meter grid lines, and the vertical illuminance uniformity ratios were
calculated along these grid lines using the same method described previously for horizontal
illuminance uniformity ratios. Figure 29 through Figure 31 show the laboratory vertical
illuminance profiles along the three grid lines. Clearly, the HPS system had a much more
concentrated vertical illuminance level in the close vicinity of the luminaire. As the distance
increases both longitudinally and transversely, the peak vertical illuminance readings for the
HPS system decreased quickly to a level similar to that of most LED systems.
Figure 29. Vertical Illuminance Along 3-Meter Grid Line
0
5
10
15
20
25
Illu
min
ance
(lu
x)
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Illu
min
ance
(lu
x)
Distance (Meter)
HPS 250 1_3 Design A 1_3 Design B 1_3 Design C 1_3 Design E 1_3 Design D 1_3
31
Figure 30. Vertical Illuminance Along 6-Meter Grid Line
Figure 31. Vertical Illuminance Along 9-Meter Grid Line
Figure 32 through Figure 34 further illustrate the vertical illuminance uniformity at the
3-, 6-, and 9-meter laboratory grid lines. From the illustrations, the uniformity ratios of the HPS
system did not seem to be particularly higher than the LED systems, which suggested that the
vertical illuminance performance of the HPS system was comparable to some LED systems. On
the other hand, the LED Design A and D (with LED optical arrays) had relatively higher
uniformity ratios.
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Illu
min
ance
(lu
x)
Distance (Meter)
HPS 250 1_6 Design A 1_6 Design B 1_6 Design C 1_6 Design E 1_6 Design D 1_6
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Illu
min
ance
(lu
x)
Distance (Meter)
HPS 250 1_9 Design A 1_9 Design B 1_9 Design C 1_9 Design D 1_9 Design E 1_9
32
Figure 32. Vertical Illuminance Uniformity at 3-Meter Grid Line
Figure 33. Vertical Illuminance Uniformity at 6-Meter Grid Line
Figure 34. Vertical Illuminance Uniformity at 9-Meter Grid Line
Light Trespass
Light trespass describes that portion of light falling out of the lighting design area,
causing light pollution. During this study, the research team measured the vertical illuminance
along the front most grid line (i.e., 13 m away from luminaire) and the rear most grid line (i.e., 6
m away from luminaire), both facing the luminaire. Figure 35 and Figure 36 illustrate the front
and rear trespass measurements for the HPS and LED systems, respectively. The figures seem to
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Average Uniformity Ratio Max Uniformity Ratio
Dis
tan
ce (
m)
HPS 250 (1) Design A (1) Design B (1)
Design C (1) Design D (1) Design E (1)
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Average Uniformity Ratio Max Uniformity Ratio
Dis
tan
ce (
m)
HPS 250 (1) Design A (1) Design B (1) Design C (1) Design D (1) Design E (1)
0.0
2.0
4.0
6.0
8.0
10.0
12.0
Average Uniformity Ratio Max Uniformity Ratio
Dis
tan
ce (
m)
HPS 250 (1) Design A (1) Design B (1) Design C (1) Design D (1) Design E (1)
33
suggest that the HPS and LED Design A and D luminaires had relatively high front and rear
trespass levels. In particular, the rear trespass readings of the HPS systems were the highest
among all evaluated luminaires.
Figure 35. Front Trespass Measured Based on VTTI Laboratory Grid
Figure 36. Rear Trespass Measured Based on VTTI Laboratory Grid
0.0
5.0
10.0
15.0
20.0
25.0
19181716151413121110 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1011121314151617181920
Illu
min
ance
(lu
x)
Design A (1) Design B (1) Design C (1)
Design D (1) Design E (1) HPS 250W (1)
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
19181716151413121110 9 8 7 6 5 4 3 2 1 0 1 2 3 4 5 6 7 8 9 1011121314151617181920
Illu
min
ance
(lu
x)
Design A (1) Design B (1) Design C (1)
Design D (1) Design E (1) HPS 250W (1)
34
Initial Laboratory Spectral Power Distribution
An SPD measurement describes the power per unit area per unit wavelength of an
illuminating body. In practice, SPD frequently refers to the graphic representation of the relative
power at each wavelength. SPD curves provide users with a visual profile of the color
characteristics of a light source and are one of the most powerful tools for determining the
spectral content of a light source. The CCT of a light source gives a good indication of its
general appearance, but does not give information on its specific spectral power distribution.
Therefore, two luminaires with similar CCTs may appear to be the same color, but their effects
on object colors can be quite different if their SPDs are significantly different.
Figure 37 shows the SPD curves for the evaluated LED and HPS systems. Note that the
visible range of light wavelengths for typical human eyes is from 390 to 750 nanometers (nm).
Within this range, the visible indigo light has a wavelength of about 445 nm, the visible yellow
light has a wavelength of about 570 nm, and the visible red light has a wavelength of about 650
nm.
Figure 37. SPD Curves for Evaluated LED and HPS Systems
As Figure 37 indicates, the HPS systems showed high relative orange and reddish light
content. In comparison, most LED systems contained relatively high relative of bluish/indigo
light content. Figure 38 includes photos of the evaluated luminaire systems with an emphasis on
their light colors.
-40
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Rel
ativ
e Ir
rad
ian
ce
Wave Length (nm)
Chart Title
Design A (1) Initial Design A (2) Initial
Design B (1) Initial Design B (2) Initial
Design C (1) Initial Design C (2) Initial
Design D (1) Initial Design D (2) Initial
Design E (1) Initial Design E (2) Initial
HPS 250W (1) Initial HPS 250W (2) Initial
35
Figure 38. Light Color of Evaluated Luminaire Systems
Initial Laboratory Power Consumption and Efficacy
Table 6 compares the actual wattages measured by the VTTI research team with the
manufacturer-rated wattages of two luminaire systems of each type. Figure 39 further illustrates
the wattage differences graphically. As shown by the illustrations, the measured wattages of
most LED systems were consistent with their rated wattages.
HPS 250
Design E
Design A Design D
Design B
Design C
36
Table 6. Luminaire Power Consumption
System Rated Wattage Measured Wattage Difference
Design A (1) 195 198.4 1.7%
Design A (2) 195 195.1 0.1%
Design B (1) 120 124.6 3.8%
Design B (2) 120 124.2 3.5%
Design C (1) 148 146.8 -0.8%
Design C (2) 148 146.0 -1.4%
Design D (1) 150 174.3 16.2%
Design D (2) 150 176.3 17.5%
Design E (1) 200 202.1 1.1%
Design E (2) 200 202.4 1.2%
HPS 250W (1)* 250 304.9 22.0%
HPS 250W (2)* 250 307.3 22.9%
*The measured HPS wattages include driver wattage.
Figure 39. Manufacturer-Rated Versus Measured (Including Driver) Wattage
Table 7 and Figure 40 illustrate the estimated luminaire efficacy based on the average
horizontal illuminance over the entire measuring grade and the measured luminaire wattage. The
results showed that the estimated efficacy values of most LED systems evaluated were generally
lower than that of the HPS systems. The exception was the Design C systems, which showed the
highest efficacy among all systems evaluated.
Figure 40. Measured Luminaire Efficacy Based on Average Horizontal Illuminance
195 195
120 120
148 148 150 150
200 200
250 250
198 195
125 124147 146
174 176
202 202
305 307
Chart Title
Rated Wattage (W) Measured Wattage (W) (includes driver)
0.06 0.06
0.08 0.080.09
0.10
0.070.08 0.08 0.08
0.09 0.08
Effi
cacy
(Lu
x/W
att)
Power Consumption (average illuminance / wattage)
37
Table 7. Measured Luminaire Efficacy based on Average Horizontal Illuminance
Luminaire
Average
Horizontal
Illuminance (lux)
Average Vertical
Illuminance (lux)
Measured
Wattage
(includes driver)
Horizontal
Illuminance Per
Wattage (lux/w)
Average
Design A (1) 12.22 14.71 198.4 0.06 0.06
Design A (2) 12.07 14.99 195.1 0.06
Design B (1) 10.09 10.28 124.6 0.08 0.08
Design B (2) 10.35 10.54 124.2 0.08
Design C (1) 13.32 12.50 146.8 0.09 0.10
Design C (2) 14.95 12.64 146 0.10
Design D (1) 12.84 14.70 174.3 0.07 0.08
Design D (2) 14.77 14.99 176.3 0.08
Design E (1) 16.78 16.90 202.1 0.08 0.08
Design E (2) 16.08 16.28 202.4 0.08
HPS 250W (1) 26.52 24.48 304.9 0.09 0.09
HPS 250W (2) 25.65 23.06 307.3 0.08
Laboratory Luminaire Performance Change Over Time
This section compares the results between first and second rounds of laboratory testing.
The second round of laboratory testing was conducted after approximately 2 years (or 8,800
hours) of field operation.
Change in Horizontal and Vertical Illuminance Over Time
Table 8 lists the average horizontal illuminance measurements for the evaluated
luminaires based on the initial and second laboratory testing, followed by Figure 41 through
Figure 43 comparing the differences graphically. To compare performance of different designs,
the research team grouped the luminaires with exposed optic array design (i.e., Design A, D, and
E). Note that the before and after testing results showed erratic performance for Design D (2)
luminaire. When comparing average performance metrics, the researchers provided values both
with and without that luminaire to account for its impact on results.
Table 8. Change in Horizontal Illuminance: Laboratory Testing
Luminaire Average Illuminance (Lux) Lumen
Depreciation
Dirt
Depreciation
Overall Light
Loss Before After (Clean) After (Dirty)
Design A (1) 11.3 10.7 10.6 -5.5% -0.8% -6.3%
Design A (2) 11.2 10.6 10.6 -5.1% 0.2% -4.9%
Design B (1) 8.8 8.4 8.4 -4.0% -1.0% -5.0%
Design B (2) 9.0 8.3 8.4 -7.5% 0.3% -7.2%
Design C (2) 12.8 12.3 11.9 -3.8% -2.7% -6.5%
Design D (1) 11.9 11.5 11.4 -3.4% -0.7% -4.2%
Design D (2) 13.2 10.4 11.5 -20.9% 7.9% -13.0%
Design E (1) 14.6 13.9 13.5 -4.9% -2.5% -7.4%
Design E (2) 14.2 13.5 13.5 -4.5% -0.2% -4.7%
HPS 250 (2) 22.5 21.3 20.3 -5.1% -4.7% -9.8%
Optic Average (A, D, E w/o D(2)) -4.7% -0.8% -5.5%
Design B Average -5.8% -0.3% -6.1%
LED Average - with D (2) -6.5% -0.4% -6.9%
LED Average - without D (2) -4.9% -1.3% -6.2%
38
Figure 41. Average Horizontal Illuminance – Initial Testing and Second Testing
Figure 42. Light Loss Based on Horizontal Illuminance
Figure 43. Comparison of Horizontal Illuminance Loss Among LED Designs
As the illustrations show, most luminaires showed light losses after the 2 years of field
operation. Without considering the outlier (Design D [2]), the results suggested an overall light
loss of 6% for the LED luminaires compared to 10% for the HPS system. When looking at the
light loss due to dirt accumulated over time, the LED systems had a light loss about 1%
compared to 5% for the HPS system. This result was confirmed during visual inspections as
well. The warm HPS luminaire attracted a significant number of insects into the housing, and a
large number of insect remains were found inside the lens. In contrast the LED systems are
much less attractive to insects due to their cooler operating temperature. Their optical
assemblies are sealed, and there was only minor dirty accumulation on the outside of their lenses.
0.0
5.0
10.0
15.0
20.0
25.0
Design A (1) Design B (1) Design C (2) Design D (1) Design E (1) HPS 250 (2)Ave
rage
Ho
rizo
nta
l Illu
min
ance
(Lu
x)
Average Illuminance Before Average Illuminance After (Clean)
Average Illuminance After (Dirty)
-10%-9%-8%-7%-6%-5%-4%-3%-2%-1%0%
Design A (1) Design B (1) Design C (2) Design D (1) Design E (1) HPS 250 (2)LED Average(with D (2))
LED Average(w/o D (2))
Chart Title
Lumen Depreciation Dirt Depreciation Overall
-10%-9%-8%-7%-6%-5%-4%-3%-2%-1%0%
Design C (2) HPS 250 (2)Optic Average (A, D, E
w/o D(2)) Design B Average
Chart Title
Lumen Depreciation
Dirt Depreciation
Overall
39
When comparing among the different LED designs, the design B luminaires seemed to
have the least light loss due to dirt depreciation while the Design C luminaire had the most light
loss. On the other hand, the Design B luminaires had the most light loss due to lumen
depreciation while Design C had the least lumen depreciation. Note that Design E (1)
experienced more dirt depreciation than Design E (2) due to its exposure to more bus traffic.
The seemingly unreasonable results of the LED Design D (2) luminaire were likely due
to its unstable light output, as the initial and clean measurements were not comparable to those of
the luminaire D (1). Other factors that might have contributed to the results include ambient
lighting, luminaire cleaning process, and to a lesser degree, issues with data collection
equipment. Note that Luminaire Design D (2) was also installed above the bus route/station in
the parking lot, which exposed more significant dirt/dust effect.
Table 9 lists the laboratory testing results for vertical illuminance, and Figure 44 through
Figure 46 further illustrate the differences between the vertical illuminance readings of the two
rounds of testing and between those of dirty and clean luminaires.
Table 9. Change in Vertical Illuminance: Laboratory Testing
Luminaire
Average
Illuminance
Before (Lux)
Average
Illuminance
After (Clean)
(Lux)
Average
Illuminance
After (Dirty)
(Lux)
Lumen
Depreciation
Dirt
Depreciation
Overall
Light Loss
Design A (1) 13.6 13.2 12.9 -2.4% -2.6% -5.0%
Design A (2) 13.6 13.2 12.6 -2.8% -4.4% -7.2%
Design B (1) 9.5 9.5 9.3 0.1% -1.5% -1.4%
Design B (2) 9.8 9.1 9.1 -7.4% -0.4% -7.9%
Design C (2) 11.4 10.5 10.2 -7.9% -2.5% -10.3%
Design D (1) 13.2 13.2 13.0 -0.2% -1.5% -1.8%
Design D (2) 13.5 11.1 12.2 -17.6% 7.9% -9.7%
Design E (1) 15.2 14.3 13.9 -6.1% -2.6% -8.7%
Design E (2) 14.8 13.3 13.2 -9.7% -0.8% -10.5%
HPS 250 (2) 21.0 21.3 20.3 1.9% -5.1% -3.2%
Optic Average (A, D, E w/o D(2)) -4.2% -2.4% -6.6%
Design B Average -3.7% -1.0% -4.6%
LED Average - with D (2) -5.2% -1.3% -6.6%
LED Average - without D (2) -3.8% -2.4% -6.2%
Figure 44. Average Vertical Illuminance – Initial Testing and Second Testing
0.0
5.0
10.0
15.0
20.0
25.0
Design A (1) Design B (1) Design C (2) Design D (1) Design E (1) HPS 250 (2)
Ave
rage
Ver
tica
l Illu
min
ance
(Lu
x)
Average Illuminance Before Average Illuminance After (Clean)
Average Illuminance After (Dirty)
40
Figure 45. Light Loss Based on Vertical Illuminance
Figure 46. Comparison of Vertical Illuminance Loss Among LED Designs
Vertical illuminance results suggested a total of 6% light loss after 2 years of operation
for the LED luminaires when accounting a more than 2% dirt depreciation, without considering
the outlier D (2) luminaire. Interestingly, the HPS system had a 2% increase in vertical
illuminance after 2 years of operation, although the vertical illuminance dirt depreciation was
about 5%, comparable to that for horizontal illuminance. This is noteworthy as the light
distribution of the luminaire is changed with a reduction in horizontal illuminance at the extents
of the light distribution can cause an increase in the vertical illuminance.
Comparing the results for vertical and horizontal illuminance measurements, the lab
testing results suggested somewhat different horizontal and vertical illuminating performance. In
terms of the LED luminaires, the dirt depreciation effect had seemingly greater impact on
vertical illuminance than on horizontal illuminance (2.4% versus 1.3%). When comparing
among the different LED designs, dirt depreciation had a less significant impact on the Design B
luminaires.
Dirt Depreciation Distribution
To better understand the impact of dirt depreciation on the luminaires, the research team
compared the light distribution of each luminaire before and after cleaning the lenses. Figure 47
through Figure 52 illustrate how dirt depreciation on horizontal illuminance for each luminaire is
-12%
-10%
-8%
-6%
-4%
-2%
0%
2%Design A (1) Design B (1) Design C (2) Design D (1) Design E (1) HPS 250 (2)
LED Average(with D (2))
LED Average(w/o D (2))
Chart Title
Lumen Depreciation Dirt Depreciation Overall
-12%
-10%
-8%
-6%
-4%
-2%
0%
2%Design C (2) HPS 250 (2)
Optic Average (A, D, Ew/o D(2)) Design B Average
Chart Title
Lumen Depreciation Dirt Depreciation Overall
41
distributed over the laboratory grid using the ratio of dirty lens illuminance to clean lens
illuminance.
The figures suggest that for the HPS luminaire, dirt depreciation primarily reduced the
horizontal illuminance of the central area and the areas on the right and left sides. This result
clearly demonstrates how dirt particles on the large lens reflected light from a central source.
The dirt depreciation impact for the Design C LED luminaires (large LED sources enclosed by a
large, flat lens) closely resembles that for the HPS luminaires, suggesting similar characteristics
between the two types of luminaires in terms of dirt depreciation. For Designs A, D, and E LED
luminaires with optic arrays, the dirt depreciation in general resulted in slightly higher horizontal
illuminance levels in the central area while lower horizontal illuminance everywhere else. This
phenomenon was found more evident for Design A, which seemed to attributable to its concave
LED optic array design. For Design B Luminaires, on the other hand, dirt depreciation resulted
in relatively uniform decrease of horizontal illuminance across the laboratory grid. Note that
Design B luminaires had large LED sources symmetrically folding towards each other, therefore
cancelling out to a certain extent the dirt reflection impact caused by each individual LED panel.
Figure 47. Dirt Depreciation Distribution - HPS 250W (2)
Figure 48. Dirt Depreciation Distribution – Design A (1)
42
Figure 49. Dirt Depreciation Distribution – Design B (1)
Figure 50. Dirt Depreciation Distribution – Design C (2)
Figure 51. Dirt Depreciation Distribution – Design D (1)
43
Figure 52. Dirt Depreciation Distribution – Design E (1)
Spectral Power Distribution
Figure 53 through Figure 58 show the SPD curves for each type of luminaires,
respectively. As the figures illustrate, the relative intensity of the yellowish light (wavelength
between 500 and 600 nm) for most LED luminaires seemingly decreased after the 2 years’
operation. According to the data, the accumulated dirt over the 2 years’ operation affected the
SPD of the LED Design C and E more. In both cases, the dirt tended to increase the output of
the yellowish light (wavelength between 500 and 600 nm) in the spectrum.
In terms of the HPS system, the second lab testing results showed that the relative output
of the reddish light (i.e., wavelength greater than 600 nm) increased after 2 years’ operation. In
addition, the accumulated dirt on the luminaire lens seemingly delayed this trend based on the
results.
Figure 53. Before and After Spectral Power Distribution – Design A
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Rel
ativ
e Ir
rad
ian
ce
Wave Length (nm)
Chart Title
Design A (1) Initial Design A (1) After Dirty Design A (1) After Clean
44
Figure 54. Before and After Spectral Power Distribution – Design B
Figure 55. Before and After Spectral Power Distribution – Design C
Figure 56. Before and After Spectral Power Distribution – Design D
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Rel
ativ
e Ir
rad
ian
ce
Wave Length (nm)
Chart Title
Design B (1) Initial Design B (1) After Dirty Design B (1) After Clean
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Rel
ativ
e Ir
rad
ian
ce
Wave Length (nm)
Chart Title
Design C (2) Initial Design C (2) After Dirty Design C (2) After Clean
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Rel
ativ
e Ir
rad
ian
ce
Wave Length (nm)
Chart Title
Design D (1) Initial Design D (1) After Dirty Design D (1) After Clean
45
Figure 57. Before and After Spectral Power Distribution – Design E
Figure 58. Before and After Spectral Power Distribution – HPS
Power Consumption and Efficacy
Figure 59 and Figure 60 compare the before and after power consumption and measured
efficacy, respectively, followed by the detailed values in Table 10. As Figure 59 suggests, most
LED luminaires did not have significant changes in power consumption after 2 years of
operation. The measured wattages for some LED luminaires even decreased. The measured
wattage of the HPS luminaire, however, increased by a non-trivial percentage after 2 years of
usage. Notice that the power consumption for the LED Design D (2) luminaire measured during
the second lab testing was significantly lower than its initial wattage. This wattage decrease
coincided with a significant decrease in light output as observed previously.
Figure 59. Initial and After Power Consumption (Including Drivers)
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Rel
ativ
e Ir
rad
ian
ce
Wave Length (nm)
Chart Title
Design E (1) Initial Design E (1) After Dirty Design E (1) After Clean
-20
0
20
40
60
80
100
120
40
0
41
0
42
0
43
0
44
0
45
0
46
0
47
0
48
0
49
0
50
0
51
0
52
0
53
0
54
0
55
0
56
0
57
0
58
0
59
0
60
0
61
0
62
0
63
0
64
0
65
0
66
0
67
0
68
0
69
0
70
0
71
0
72
0
73
0
74
0
75
0
76
0
77
0
78
0
79
0
80
0
Re
lati
ve Ir
rad
ian
ce
Wave Length (nm)
Chart Title
HPS 250W (2) Initial HPS 250W (2) After Dirty HPS 250W (2) After Clean
198 195
125 124146
174 176202 202
307
203 204
130 129146
174159
196 197
330
Design A(1)
Design A(2)
Design B(1)
Design B(2)
Design C(2)
Design D(1)
Design D(2)
Design E(1)
Design E(2)
HPS250W (2)
Wat
tage
(W
)
Chart Title
Initial Wattage Wattage after Operation
46
Figure 60. Initial and After Efficacy
Table 10. Lab Measured Before and After Power Consumption and Efficacy
Luminaire Rated
Wattage
Initial
Wattage
Wattage after
Operation
Initial
Efficacy
After Efficacy
(Dirty)
After Efficacy
(Clean)
Design A (1) 195 198.4 203.4 2.5% 0.06 0.05 -15.4% 0.05 -14.7%
Design A (2) 195 195.1 204.0 4.6% 0.06 0.05 -15.7% 0.05 -15.9%
Design B (1) 120 124.6 129.7 4.1% 0.08 0.06 -20.4% 0.07 -19.6%
Design B (2) 120 124.2 129.4 4.2% 0.08 0.06 -22.5% 0.06 -22.8%
Design C (2) 148 146.0 146.3 0.2% 0.10 0.08 -20.4% 0.08 -18.1%
Design D (1) 150 174.3 174.4 0.1% 0.07 0.07 -11.0% 0.07 -10.3%
Design D (2) 150 176.3 158.7 -10.0% 0.08 0.07 -13.8% 0.07 -21.6%
Design E (1) 200 202.1 195.8 -3.1% 0.08 0.07 -16.9% 0.07 -14.6%
Design E (2) 200 202.4 196.8 -2.8% 0.08 0.07 -13.6% 0.07 -13.4%
LED Average 0.0% 0.08 0.07 -16.6% 0.07 -16.8%
HPS 250W (2) 250 307.3 329.7 7.3% 0.08 0.06 -26.3% 0.06 -22.4%
When combining changes in horizontal illuminance and wattages, the calculated
luminaire efficacies for most luminaires decreased after 2 years of operation. When comparing
between HPS and LED luminaires, the data suggested that the efficacy of the HPS luminaire
decreased almost 10% more than that of LED luminaires without cleaning their lenses.
Field Horizontal Illuminance
The research team took horizontal illuminance measurements over the field grid during
each site visit. Table 11 and Figure 61 illustrate the mean horizontal illuminance values over the
field grid for all light systems. Figure 62 through Figure 67 compare the field horizontal
illuminance values of different luminaires along the 0-meter grid line (i.e., directly under the
luminaires. Figure 68 shows the changes of light output for the evaluated LED systems based on
the field testing data in an effort to understand the overall light loss factor of the LED
technology. All readings shown in the illustrations have been corrected for ambient lighting and
temperature impacts. Note that this analysis focused on the illuminance changes over time of
different LED designs. The absolute illuminance values were based on a custom grid that may
not necessarily meet lighting design requirements.
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Design A(1)
Design A(2)
Design B(1)
Design B(2)
Design C(2)
Design D(1)
Design D(2)
Design E(1)
Design E(2)
HPS 250W(2)
Effi
cacy
(Lu
x/W
att)
Chart Title
Initial Efficacy After Efficacy (Dirty) After Efficacy (Clean)
47
Table 11. Mean Horizontal Illuminance Measurements on Field Grid
Date Mean Horizontal Illuminance (Lux) and Percent Illuminance Change Against 2012/09 Value
Design A (1) Design B (1) Design C (1) Design D (1) Design E (1) HPS 250W
2012/09 9.4 0% 9.17 0% 16.23 0% 11.97 0% 13.93 0% 23.63 0%
2012/12 10.17 8.2% 9.83 7.2% 16.89 4.1% 12.27 2.5% 14.12 1.4% 25.45 7.7%
2013/03 9.77 3.9% 9.19 0.2% 16.11 -0.7% 10.58 -11.6% 13.2 -5.2% 27.33 15.7%
2013/06 9.67 2.9% 9.48 3.4% 16.08 -0.9% 12.05 0.7% 13.48 -3.2% 26.56 12.4%
2013/09 9.37 -0.3% 10.12 10.4% 16.1 -0.8% 11.38 -4.9% 13.41 -3.7% n/a n/a
2013/12 10.5 11.7% 10.13 10.5% 17.22 6.1% 10.41 -13.0% 14.68 5.4% 27.48 16.3%
2014/03 10.18 8.3% 9.85 7.4% 16.94 4.4% 10.65 -11.0% 13.04 -6.4% 28.34 19.9%
2014/06 8.29 -11.8% 8.35 -8.9% 13.78 -15.1% 10.9 -8.9% 12.02 -13.7% 24.88 5.3%
2014/09 9.27 -1.4% 8.89 -3.1% 15.26 -6.0% 11.78 -1.6% 13.76 -1.2% 26.71 13.0%
Figure 61. Average Horizontal Illuminance on Field Grid
Figure 62. Field Horizontal Illuminance Along 0-Meter Grid Line – Design A (1)
Figure 63. Field Horizontal Illuminance Along 0-Meter Grid Line – Design B (1)
0
5
10
15
20
25
30
Design A (1) Design B (1) Design E (1) Design D (1) Design C (1) HPS 250W
Illu
min
ance
(lu
x)
Average Horizontal Illuminance
2012_09 2012_12 2013_03 2013_06 2013_09
2013_12 2014_03 2014_06 2014_09
0
2
4
6
8
10
12
14
16
18 15 12 9 6 3 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_09
2013_12 2014_03 2014_06 2014_09
0
2
4
6
8
10
12
14
16
18
18 15 12 9 6 3 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_09
2013_12 2014_03 2014_06 2014_09
48
Figure 64. Field Horizontal Illuminance Along 0-Meter Grid Line – Design C (1)
Figure 65. Field Horizontal Illuminance Along 0-Meter Grid Line – Design D (1)
Figure 66. Field Horizontal Illuminance Along 0-Meter Grid Line – Design E (1)
0
5
10
15
20
25
30
35
40
18 15 12 9 6 3 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03
2013_06 2013_09 2013_12
2014_03 2014_06 2014_09
0
5
10
15
20
25
18 15 12 9 6 3 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_09
2013_12 2014_03 2014_06 2014_09
0
5
10
15
20
25
30
35
40
18 15 12 9 6 3 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03
2013_06 2013_09 2013_12
2014_06 2014_09
49
Figure 67. Field Horizontal Illuminance Along 0-Meter Grid Line – HPS 250W
Figure 68. Changes in Light Output for LED Luminaires During Field Testing
Based on the field horizontal illuminance data, the researchers could not conclude a clear
trend in significant light reduction over the 2-year field evaluation period. Note that Figure 62
and Figure 63 suggest that horizontal light distribution of the LED Design A and B luminaires
changed over time in the field. This phenomenon might be due to tilt/roll of the luminaires over
time and/or irregular changes of the performance of luminaire LED optics.
One aspect that impacted the horizontal illuminance readings (including their uniformity
and symmetry) of the luminaires was how the luminaire was initially installed in terms of its
orientation and tilt (levelness). This factor also affected the vertical illuminance results as
discussed in the following sections. Due to human factors and available lighting fixture
characteristics, not all luminaires were installed perfectly level. Figure 69 shows the orientation
of the evaluated luminaires that can be used to better understand the horizontal illuminance
measurements. Notice that the luminaire found to be installed in the least level orientation was
the Design A (1) luminaire (Figure 70).
0
10
20
30
40
50
60
18 15 12 9 6 3 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03
2013_06 2013_12 2014_06
2014_09
85%
90%
95%
100%
105%
110%
115%
120%
2012_09 2012_12 2013_03 2013_06 2013_09 2013_12 2014_03 2014_06 2014_09
Per
cen
t o
f In
itia
l Illu
min
ance
Design A (1) Design B (1) Design E (1) Design D (1) Design C (1) Average
50
Tilt (°) Roll (°)
Design A (1) 9 2
Design B (1) 2 1
Design C (1) 0 0
Design D (1) 0 0
Design E (1) 0 6
Figure 69. Luminaire Orientation Measurements
Figure 70. Field Luminaire Orientation – Design A (1)
Field Vertical Illuminance
Table 12 and Figure 71 show the average vertical illuminance values along the field grid
used for evaluation. Figure 72 through Figure 77 further illustrate the vertical illuminance values
along the 0-meter grid line in the field. The figures suggest that, other than Design B, most LED
luminaires in general exhibited decreasing vertical illuminance levels over time. As Table 12
shows, the overall vertical illuminance reduction during the 2 years of operation was around 10%
for most LED luminaires. The only exception for this trend was the Design B luminaire, which
had a three-panel folding design with the central panel facing directly downwards and the other
two folding towards the center. These observations may suggest that dirt depreciation has a
more significant impact on vertical illuminance than on horizontal illuminance, which is
consistent with the lab testing results.
Field vertical illuminance analysis did not reveal noticeable changes in vertical
illuminance distribution. Figure 72 through Figure 77 also suggested that the vertical
51
illuminance distribution along the 0-meter grid line (i.e., underneath the luminaire) maintained
the same pattern in most cases over the 2 years.
Table 12. Mean Vertical Illuminance Measurements on Field Grid
Date Mean Vertical Illuminance on Field Grid (Lux)
Design A (1) Design B (1) Design C (1) Design D (1) Design E (1) HPS 250W
2012/09 9.5 8.6 11.6 12.9 12.6 17.2
2012/12 10.1 8.9 11.1 13.0 12.3 17.3
2013/03 9.9 8.7 11.3 11.5 11.7 17.8
2013/06 9.5 9.1 11.2 12.6 12.1 18.6
2013/09 9.5 9.8 11.1 12.3 12.1 -
2013/12 9.6 8.6 11.6 11.0 12.0 16.1
2014/03 9.6 8.7 11.3 10.7 11.0 17.4
2014/06 8.5 8.3 10.4 11.2 11.6 16.7
2014/09 8.6 8.3 10.3 11.4 11.5 17.3
2-Year Change -10.1% -3.5% -10.8% -12.0% -8.2% 0.3%
Figure 71. Mean Vertical Illuminance on Field Grid
Figure 72. Field Vertical Illuminance Along 0-Meter Grid Line – Design A (1)
0
2
4
6
8
10
12
14
16
18
20
Design A (1) Design B (1) Design C (1) Design D (1) Design E (1) HPS 250W
Illu
min
ance
(lu
x)
Average Vertical Illuminance
2012_09 2012_12 2013_03 2013_06 2013_09
2013_12 2014_03 2014_06 2014_09
0
2
4
6
8
10
12
14
16
18
20
18 15 12 9 6 3 0 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03
2013_06 2013_09 2013_12
2014_03 2014_06 2014_09
52
Figure 73. Field Vertical Illuminance Along 0-Meter Grid Line – Design B (1)
Figure 74. Field Vertical Illuminance Along 0-Meter Grid Line – Design C (1)
Figure 75. Field Vertical Illuminance Along 0-Meter Grid Line – Design D (1)
0
2
4
6
8
10
12
14
16
18 15 12 9 6 3 0 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12
2013_03 2013_06
2013_09 2013_12
2014_03 2014_06
2014_09
0
5
10
15
20
25
18 15 12 9 6 3 0 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_09 2013_12 2014_03 2014_06 2014_09
0
5
10
15
20
25
30
18 15 12 9 6 3 0 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_09
2013_12 2014_03 2014_06 2014_09
53
Figure 76. Field Vertical Illuminance Along 0-Meter Grid Line – Design E (1)
Figure 77. Field Vertical Illuminance Along 0-Meter Grid Line – HPS 250W
Correlated Color Temperature
The color temperature of the electromagnetic radiation emitted from an ideal black body
is defined as its surface temperature in Kelvin. LEDs emit light primarily by processes other
than thermal radiation and the emitted radiation does not follow the form of a black-body
spectrum. The lighting community uses CCT as a standard for comparing the color of LED light
sources. CCT is the color temperature of a black-body radiator which to human color perception
most closely matches the subject LED light.
Table 13 and Figure 78 compare the manufacturer-rated CCTs with the CCTs measured
during the field evaluation. It is important to note that the Design B luminaires had a higher
rated CCT than the other designs. The data showed that most LED luminaire systems exhibited
CCTs consistent with the manufacturer specifications. The Design B luminaire was the only one
that resulted in measured CCTs not consistent with the specification. It is important to note that
LEDs with higher CCT typically have a higher light output than those with a lower CCT due to
the efficiency of the phosphor in the LED chip. Typically, a 6500K luminaire would not be
compared to a 4100 K luminaire.
0
5
10
15
20
25
30
18 15 12 9 6 3 0 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_09 2013_12 2014_03 2014_06 2014_09
0
5
10
15
20
25
30
18 15 12 9 6 3 0 0 3 6 9 12 15 18
Illu
min
ance
(lu
x)
Distance (m)
2012_09 2012_12 2013_03 2013_06 2013_12 2014_06 2014_09
54
The field measurements also seemed to suggest that most LED luminaires had
moderately decreasing CCTs and become slightly more yellowish/reddish over time (Figure 79)
over time.
Table 13. Specified and Measured CCTs for Evaluated LED Luminaires
Luminaire Specification
CCT (K) 12-Sep 13-Jun 13-Sep 13-Dec 14-Mar 14-Jun 14-Sep
2-Year
Change
Design A (1) 4300 4331 4276 4295 4253 4188 4159 4091 -5.5%
Design B (1) 5000+/-300 6501 5881 6404 6381 6148 6169 5974 -8.1%
Design C (1) 4000 4394 4427 4549 4325 4352 4373 4306 -2.0%
Design D (1) 4000 4135 4080 4108 4076 4035 4053 4012 -3.0%
Design E (1) 4000 4390 4355 4368 4323 4342 4302 4206 -4.2%
Figure 78. Specified and Measured CCTs for Evaluated LED Luminaire Systems
Figure 79. CCT Trendlines (Polynomial) During First 2 Years of Operation
0
1000
2000
3000
4000
5000
6000
7000
Design A (1) Design B (1) Design C (1) Design D (1) Design E (1)
CC
T (K
)
Correlated Color Temperature (CCT) comparison to specification
Specification CCT (K) Measured CCT (K) 12-Sep Measured CCT (K) 13-JunMeasured CCT (K) 13-Sep Measured CCT (K) 13-Dec Measured CCT (K) 14-MarMeasured CCT (K) 14-Jun Measured CCT (K) 14-Sep Linear (Measured CCT (K) 14-Jun)
55
Field Luminaire Inspection
During each field trip, the research team also conducted a thorough inspection of the
evaluated luminaires. The field inspection of the luminaires included visual inspection of
exterior and interior conditions of the luminaire housing and measuring the luminaire
temperature.
Visual Inspection
Upon visual inspection of the luminaires beginning in the December 2012 data collection,
comments were made regarding dirt buildup as well as other factors (these are the same
luminaires that were selected for the detailed photometric analysis, not all of the luminaires). A
summary of the visual inspection findings is shown in Table 14.
Table 14. Visual Inspection Comments
Category Design A
(1)
Design B
(1)
Design
C (1)
Design D
(1)
Design E
(1) HPS 250W
Wildlife intrusion device installed No Yes No No Yes No
Level of presence of wildlife (e.g., Insects) Low High Medium Medium Medium Low Level of rust in component housing Low Low High Low Medium Low Level of dirt inside component housing Low Low Medium Low Low Low Level of dirt buildup on optics cover Low Low Medium Low Low High
Level of damage to electrical components Low Low Low Low Low Low
The visual inspection suggested that a number of luminaire systems did not have a
wildlife intrusion device installed. Over time, some of the evaluated luminaire exhibited dirt
accumulation and/or wildlife intrusion both in and out of the luminaire housing units. It should
be noted that even with luminaires with intrusion devices, wildlife and dirt were still found in the
luminaires. Figure 80 shows examples of such issues identified during field visual inspections.
56
Figure 80. Examples of Lack of Ingress Protection on Housing – Design C. A: Large opening around tenon;
B: Rust on housing door; C: Dirt/wildlife on housing door; D: Rust/wildlife in housing.
Cost-Benefit Analysis Results
Table 15 and Figure 81 show the total annual and accumulative lighting maintenance and
operational costs predicted for a 25-year horizon for all scenarios, including a base scenario that
assumed no HPS systems would be converted to LED in the next 25 years. The results suggested
that the investment would be returned (compared to the base scenario) in the 8th year for all
scenarios, although scenario S2 (the existing HPS systems are converted over 5 years)
corresponded with the shortest investment return time (as noted in bold in Table 15). Over the
25-year analysis period, the return-on-investment (ROI) for the different scenarios ranged
between 3.25 and 5.76, with 3.25 projected for scenario 1; 4.45 for the 5-year replacement
scenario; and 5.76 for the 10-year replacement scenario. ROI in this context is defined as the
ratio between the total accumulative cost savings over the 25-year period and the cost of
purchasing LED luminaires.
A B
C D
57
Table 15. Total Accumulative Lighting Costs for All Scenarios ($millions, 2015)
Year
Total Accumulative
Cost-Base (only HPS)
(based on Eq. 1 - 3)
Total Accumulative
Cost-S1
(based on Eq. 4 – 7)
Total Accumulative
Cost-S2
(based on Eq. 8 – 11)
Total Accumulative
Cost-S3
(based on Eq. 12 – 15)
1 21.62 55.15 28.00 24.61
2 43.68 71.97 54.34 48.60
3 66.17 89.13 79.13 72.03
4 89.12 106.64 102.49 94.97
5 112.52 124.49 124.49 117.45
6 136.39 142.70 142.48 139.51
7 160.74 161.28 160.83 161.16
8 185.58 180.22 179.54 182.41
9 210.91 199.55 198.64 203.27
10 236.75 219.26 218.11 223.74
11 263.11 239.36 237.97 243.37
12 289.99 259.87 258.23 263.40
13 317.42 280.79 278.90 283.82
14 345.39 302.13 299.98 304.66
15 373.92 323.89 321.48 325.91
16 403.02 346.09 343.41 347.58
17 432.70 368.73 365.78 369.69
18 462.97 391.83 388.59 392.25
19 493.85 415.38 411.87 415.25
20 525.35 439.41 435.61 438.71
21 557.48 463.92 459.82 462.64
22 590.25 488.92 484.52 487.05
23 623.68 514.42 509.71 511.95
24 657.78 540.43 535.40 537.35
25 692.55 566.96 561.61 563.26
Figure 81. Total Annual and Accumulative Lighting Costs for All Scenarios (2015 $)
Table 16 and Figure 82 show the annualized and total electricity costs for all scenarios.
As the illustrations show, over the 25-year analysis period, scenario S1 (replacing HPS systems
at once) would reduce the lighting energy cost by $56.6 million (48%), scenario S2 would cut the
0
100
200
300
400
500
600
700
800
0
10
20
30
40
50
60
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Tota
l Acc
um
ula
tive
Co
st (
$1
,00
0,0
00
)
Tota
l An
nu
al C
ost
($
1,0
00
,00
0)
Year
Chart Title
Total Annual Cost-Base
Total Annual Cost-S1
Total Annual Cost-S2
Total Annual Cost-S3
Total AccumulativeCost-Base
Total AccumulativeCost-S1
Total AccumulativeCost-S2
Total AccumulativeCost-S3
58
energy cost by half ($58.8 million), and scenario S3 would reduce by $58.1 million (49%)
compared with the scenario where no LED technology would be used.
Table 16. Annual and Accumulative Electricity Cost ($millions, 2015)
Year
Energy Consumption
(Base, Eq. 1)
Energy Consumption
(S1, Eq. 5)
Energy Consumption
(S2, Eq. 9)
Energy Consumption
(S3, Eq. 13)
Annual Accumulative Annual Accumulative Annual Accumulative Annual Accumulative
1 3.69 3.69 1.93 1.93 3.34 3.34 3.52 3.52
2 3.77 7.46 1.97 3.89 3.02 6.37 3.40 6.91
3 3.84 11.30 2.01 5.90 2.67 9.04 3.26 10.17
4 3.92 15.22 2.05 7.94 2.29 11.33 3.10 13.28
5 4.00 19.22 2.09 10.03 1.87 13.20 2.93 16.21
6 4.08 23.30 2.13 12.16 1.91 15.11 2.75 18.96
7 4.16 27.46 2.17 14.33 1.95 17.05 2.55 21.50
8 4.24 31.70 2.21 16.54 1.99 19.04 2.33 23.83
9 4.33 36.03 2.26 18.80 2.02 21.06 2.09 25.92
10 4.41 40.45 2.30 21.10 2.07 23.13 1.84 27.76
11 4.50 44.95 2.35 23.45 2.11 25.24 1.88 29.64
12 4.59 49.54 2.40 25.85 2.15 27.39 1.91 31.55
13 4.68 54.23 2.44 28.29 2.19 29.58 1.95 33.50
14 4.78 59.01 2.49 30.79 2.24 31.81 1.99 35.50
15 4.87 63.88 2.54 33.33 2.28 34.09 2.03 37.53
16 4.97 68.85 2.59 35.92 2.33 36.42 2.07 39.60
17 5.07 73.92 2.65 38.57 2.37 38.79 2.11 41.71
18 5.17 79.09 2.70 41.27 2.42 41.21 2.16 43.87
19 5.28 84.37 2.75 44.02 2.47 43.68 2.20 46.06
20 5.38 89.75 2.81 46.83 2.52 46.20 2.24 48.31
21 5.49 95.24 2.86 49.69 2.57 48.77 2.29 50.59
22 5.60 100.84 2.92 52.61 2.62 51.39 2.33 52.93
23 5.71 106.55 2.98 55.59 2.67 54.06 2.38 55.31
24 5.82 112.38 3.04 58.63 2.73 56.78 2.43 57.73
25 5.94 118.32 3.10 61.73 2.78 59.56 2.48 60.21
Figure 82. Annual and Accumulative Energy Consumption for All Scenarios (2015 $)
Table 17 and Figure 83 further show the luminaire and ancillary structure maintenance
costs (including HPS lamp replacement costs) required for all scenarios. According to the
analysis, replacing all HPS luminaires at the first year would result in a 19% saving in
0
20
40
60
80
100
120
0
1
2
3
4
5
6
7
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Acc
um
ula
tive
En
ergy
Co
st (
$1
,00
0,0
00
)
An
nu
al E
ner
gy C
ost
($
1,0
00
,00
0)
Year
Chart Title
Annual EnergyCost-Base
Annual EnergyCost-S1
Annual EnergyCost-S2
Annual EnergyCost-S3
AccumulativeEnergy Cost-Base
AccumulativeEnergy Cost-S1
AccumulativeEnergy Cost-S2
AccumulativeEnergy Cost-S3
59
maintenance and related costs, the 5-year scenario would result in an 18% saving, and the 10-
year scenario corresponded to a 16% saving in maintenance costs. LED luminaires require less
maintenance over time (arguably maintenance free) due to their much longer service lives.
Table 17. Annual and Accumulative Maintenance Cost ($millions, 2015)
Year
Maintenance Cost
(Base, Eq. 2)
Maintenance Cost
(S1, Eq. 6)
Maintenance Cost
(S2, Eq. 10)
Maintenance Cost
(S3, Eq. 14)
Annual Accumulative Annual Accumulative Annual Accumulative Annual Accumulative
1 17.93 17.93 14.57 14.57 16.93 16.93 17.23 17.23
2 18.29 36.21 14.86 29.42 16.67 33.60 17.27 34.50
3 18.65 54.87 15.15 44.58 16.39 49.99 17.31 51.81
4 19.03 73.89 15.46 60.04 16.09 66.07 17.34 69.15
5 19.41 93.30 15.77 75.80 15.77 81.84 17.37 86.52
6 19.79 113.09 16.08 91.89 16.08 97.92 17.39 103.91
7 20.19 133.28 16.40 108.29 16.40 114.33 17.40 121.31
8 20.59 153.87 16.73 125.02 16.73 131.06 17.41 138.72
9 21.01 174.88 17.07 142.09 17.07 148.13 17.41 156.14
10 21.43 196.31 17.41 159.50 17.41 165.53 17.41 173.55
11 21.85 218.16 17.76 177.25 17.76 183.29 17.76 191.30
12 22.29 240.45 18.11 195.37 18.11 201.40 18.11 209.41
13 22.74 263.19 18.47 213.84 18.47 219.88 18.47 227.89
14 23.19 286.38 18.84 232.68 18.84 238.72 18.84 246.73
15 23.66 310.04 19.22 251.90 19.22 257.94 19.22 265.95
16 24.13 334.16 19.60 271.51 19.60 277.54 19.60 285.56
17 24.61 358.78 20.00 291.50 20.00 297.54 20.00 305.55
18 25.10 383.88 20.40 311.90 20.40 317.94 20.40 325.95
19 25.61 409.48 20.80 332.71 20.80 338.74 20.80 346.75
20 26.12 435.60 21.22 353.93 21.22 359.96 21.22 367.97
21 26.64 462.24 21.64 375.57 21.64 381.61 21.64 389.62
22 27.17 489.41 22.08 397.65 22.08 403.68 22.08 411.70
23 27.72 517.13 22.52 420.17 22.52 426.20 22.52 434.22
24 28.27 545.40 22.97 443.14 22.97 449.17 22.97 457.19
25 28.84 574.24 23.43 466.57 23.43 472.60 23.43 480.61
Figure 83. Annual and Accumulative Maintenance Cost for All Scenarios (2015 $)
0
100
200
300
400
500
600
0
5
10
15
20
25
30
35
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Acc
um
ula
tive
Mai
nte
nan
ce C
ost
($
1,0
00
,00
0)
An
nu
al M
ain
ten
ance
Co
st (
$1
,00
0,0
00
)
Year
Chart Title
Annual MaintenanceCost-Base
Annual MaintenanceCost-S1
Annual MaintenanceCost-S2
Annual MaintenanceCost-S3
AccumulativeMaintenance Cost-Base
AccumulativeMaintenance Cost-S1
AccumulativeMaintenance Cost-S2
AccumulativeMaintenance Cost-S3
60
DISCUSSION
Discussion of Laboratory Testing Results
The following discusses the findings based on the laboratory testing results:
Horizontal illuminance. Laboratory testing results suggested that the LED luminaires
evaluated provided lower levels of horizontal illuminance over the VTTI laboratory
grid compared to the HPS luminaires. However, light output of the HPS systems
concentrated in a relatively limited area, suggesting poor performance in achieving
light uniformity. Among the different LED designs, the Design C LED luminaires
contained large LED light sources and therefore resembled the horizontal illuminance
distribution of the HPS systems more than other LED designs. Overall, the different
LED designs exhibited very different horizontal illuminance distribution patterns,
with Design C and E more elongated while Design A, B, and D more circular. The
different patterns suggest that a clear understanding of the horizontal illuminance
pattern of a LED luminaire is critical for developing the most cost-effective lighting
design of a specific application.
Vertical illuminance. The evaluated HPS luminaires provided a higher level of
average vertical illuminance over the VTTI laboratory grid than the LED luminaires.
In addition, the HPS systems provided highly concentrated vertical illuminance levels
within the close vicinity of the luminaire, resulting in relatively intensive glare and
poor vertical illuminance uniformity. In contrast, most LED designs (e.g., Design A,
B, D, and E to a less extent) provided a much more uniform distribution of vertical
illuminance. In addition, the rear trespass levels of LED luminaires (along the
outmost grid line behind the luminaire) was all found to be lower than the HPS
luminaires.
Spectral power distribution and light quality. All LED luminaires evaluated emitted
light that is much closer to natural light in color. The HPS luminaires emitted
yellowish lights with high special power within the 490 – 510 nm (green) and 560 –
620 nm (yellow) ranges. It is widely recognized that whiter, or more naturally
colored light help drivers to better discern objects on roadways compared to
traditional light. Among the different LED designs, Design D contained the most
yellow light distribution while Design B contained the lowest yellow light.
Power consumption and efficacy. The measured wattages of most LED luminaires
were consistent with manufacturer ratings with the only exception of Design D LED
luminaires. The measured efficacy of most LED luminaires is comparable to that of
the HPS systems, with Design C luminaires slightly exceeding the HPS systems.
Light loss and light quality deterioration over time. Based on before- and after-
operation laboratory testing results, the LED luminaires had a 6% overall reduction in
light output after 2 years due to light loss and dirt depreciation, compared to the 10%
(for horizontal illuminance) or 3% (for vertical illuminance) reduction of the HPS
61
systems. The testing results suggested a much more significant light loss due to
lumen depreciation than due to dirt depreciation for LED luminaires. In addition, the
evaluated LED luminaires generally had much less dirt depreciation than did the HPS
systems. Interestingly, the lab results seemed to suggest that lumen depreciation was
more obvious in the form of horizontal illuminance than in the form of vertical
illuminance for LED luminaires. In contrast, the light loss attributable to dirt was
more significant in the form of vertical illuminance than horizontal illuminance.
Dirt depreciation distribution. The laboratory testing results suggested that, for the
HPS luminaire, dirt depreciation primarily reduced the horizontal illuminance of the
central area and the areas on the right and left sides. The dirt depreciation impact for
the Design C LED luminaires with large LED sources closely resembled that for the
HPS luminaires, suggesting similar characteristics between the two types of
luminaires in terms of dirt depreciation. For Designs A, D, and E LED luminaires
with optic arrays, the dirt depreciation in general resulted in slightly higher horizontal
illuminance levels in the central area while lower horizontal illuminance everywhere
else. For Design B Luminaires, on the other hand, dirt depreciation resulted in
relatively uniform decrease of horizontal illuminance across the laboratory grid due to
its unique three-panel folding design that evened out the dirt reflection effect.
Technological comparison. Overall, the LED systems exhibited much whiter light
output, better light uniformity, and lower glare and backlight. Among the different
LED designs, luminaires with LED optic arrays (e.g., Design A, B, and D) had better
uniformity than luminaires with large LED optics (e.g., Design C). Exposed LED
optic arrays did not attract more dirt than those covered with larger lenses, although it
would be more difficult to clean individual optics when they get dirty. Dirt
depreciation analysis results suggested that dirt accumulation on luminaires with large
light sources tend to reduce their horizontal illuminance level in the central area and
on right and left side of the light grid. From this perspective, the three-folding panel
design and/or the use of optic arrays will result in less significant impact from dirt
depreciation over time.
Discussion of Field Testing Results
The following discusses the findings based on the field testing results:
Field horizontal illuminance. Based on the field measurements, the research team
could not conclude a clear trend in significant reduction of light output during the first
2 years of field operation. When comparing with the HPS systems, the LED
luminaires in general exhibited more stable performance (i.e., less significant
performance variations over time) than the HPS systems. Among the LED designs,
the research team observed changes in horizontal light distribution over time, which
might be due to tilting/rotation of the luminaires over time and/or irregular changes of
the performance of individual LED packages.
62
Field vertical illuminance. Although not significantly, the field vertical illuminance
measurements suggested a decreasing trend for most LED designs. The overall
vertical illuminance reduction during the 2 years of operation was around 10% for
most LED luminaires. The only exception for this trend was the Design B luminaire,
which had a three-panel folding design with the central panel facing directly
downwards and the other two folding towards the center. These observations,
combined with the horizontal illuminance results, seem to suggest that dirt
depreciation has a more significant impact on vertical illuminance than on horizontal
illuminance. In addition, field vertical illuminance analysis did not reveal noticeable
changes in vertical illuminance distribution, which seems to indicate that the changes
in horizontal illuminance distribution for Design A and B were due to the
performance of LED optics instead of tilting of the luminaires.
LED luminaire light color. The study showed that CCTs measured on the field for
most LED designs were generally consistent with the manufacturer-rated CCTs. The
Design B luminaire with the highest manufacturer-rated CCT was the only one that
resulted in measured CCTs not consistent with the specified CCT. The field
measurements also suggested that most LED luminaires had a moderately decreasing
trend in CCTs over time, which indicates that the colors of the evaluated LED lights
would become slightly yellowish/reddish over time. The 2-year reduction in
measured CCTs for all LED luminaires was found to be between 2% and 8%.
Visual inspection results. Data analysis and visual inspection results suggested that,
without proper installation procedures and/or strict measurements, luminaires could
fail to meet tilt and roll requirements during installations. In addition, installed
luminaires may rotate/tilt over time when they are not securely mounted. Many
evaluated luminaires did not have sufficient ingress protection for the housing,
resulting in significant rust/dirt accumulation and wildlife intrusion inside the
electrical compartment. The researchers did not find issues with the optical assembly
ingress protection for any of the LED designs. In contrast, the HPS systems had
significant dirt and wildlife accumulation inside the optical lens.
Ambient factors affecting field study results. The field study took place over a
relatively significant duration of time. During the different field visits, a large
number of ambient factors, such as moon light, presence and change over time of
vegetation (e.g., tree leaves and grass), ambient temperature, and artificial light
(although minimal) all played a role in the lighting measurement results. Although
the research team devoted significant effort to control such factors, their impact
seemed to have affected the field study results.
Discussion of Life Cycle Cost Analysis
The results suggested that it is important to start utilizing LED technology to save energy.
However, because LED technology is expected to continually improve, the greatest benefit is
realized when LED systems are implemented gradually and systematically, which is shown in
63
the results of different scenarios taking into account the expected improvements in LED
luminaire price and efficacy. Although the research team collected a significant amount of data,
they also had to make a number of assumptions to perform the cost-benefit analysis:
Electricity cost. Currently, VDOT pays roadway lighting electricity fees both by
actual usage (i.e., kWh) and by flat fees calculated based on service areas. As such, it
was not feasible to obtain accurate rates of VDOT electricity expenditures for
roadway lighting. In addition, to materialize the energy savings when LED
luminaires are used, it is important to phase out the flat fee approach and use
electricity meters for all LED luminaires.
Implementation scenarios. In this analysis, the research team assumed three LED
implementation scenarios: replacing all HPS luminaires at once, in even increments
over 5 years, and in even increments over 10 years. If the HSP luminaires are to be
replaced on a project-by-project basis, the cost-benefit results most applicable to
VDOT will be those for the second and third scenarios: replacing all traditional
luminaires within 5 to10 years. However, VDOT is currently going through a
comprehensive energy audit that may result in much faster replacement of LED
luminaires. If this is the case, scenario S1 and, to a certain extent, S2 will be more
applicable.
Other assumptions. The research team made reasonable assumptions on several
factors, such as future LED technology improvement and price decrease, inflation
rate, and lighting maintenance needs. The results may change depending on the
accuracy of those assumptions.
LED Implementation Implications and Needs
To facilitate the implementation of LED luminaires and their long-term maintenance,
there are several strategies that can be beneficial:
Establish a LED luminaire prequalification and testing program. With the recent
advance of the LED industry, there have been a significant number of LED roadway
lighting products developed by manufacturers of different sizes and qualifications.
For each lighting project, there can be potentially a large number of products
submitted for bidding. Understanding and testing these products requires significant
expertise and in many cases is considerably time-consuming. To ensure that only
suitable and reliable products are submitted for bidding, an LED prequalification
program should be established to identify and characterize existing LED products
meeting VDOT requirements. Such a program would reduce the technical and
liability burden for product selection by contractors and project managers within tight
project schedules, and therefore reduce project delays. It would also minimize the
possibility of using faulty products for VDOT projects by allowing more expertise in
product selection and more time for thorough product testing.
64
Continuously monitor the performance of different LED luminaires to better
understand and utilize the technology. This research was based on results from a 2-
year study, which did not allow sufficient analysis of LED performance change over
time. After VDOT starts to implement LED luminaires, it will provide a great
opportunity for VDOT to continuously track the performance of the LED luminaires
for a much longer period of time. Such continuous performance monitoring will
provide critical knowledge for VDOT to update the LED specification and lighting
design standards/processes. The research team recommends that VDOT continuously
monitor the field performance of LED luminaires for at least 10 years.
Consider an on-call technical support group for LED-related issues. The vast array
of LED products with varying design and performance has challenged agencies in
ways not encountered with traditional lighting technologies. Identifying reliable and
cost effective LED products suitable for a specific application now requires more
technological know-how and product testing. Agencies may experience times when
in-house expertise cannot meet tasks such as evaluating LED lab testing reports and
technical cut sheets that are frequently not standardized within the industry,
understanding light distribution characteristics and their implications to specific
lighting designs, and identifying products with suitable/reliable photometric
performance. An on-call technical support group with reputable expertise in critical
LED and related subject areas therefore becomes beneficial to aid VDOT designers
and engineers in LED luminaire selection and testing.
Develop LED specifications for high-mast, sign, and tunnel/under-bridge lighting.
This study primarily studied LED luminaires for traditional roadway lighting. Other
roadway lighting applications, such as high-mast lighting, sign lighting, and tunnel
lighting have different requirements. The potential of using LED technology for such
applications need to be studied and specifications applicable to such applications
should be developed as well.
Phase out unmetered electrical services. Currently, VDOT still maintains unmetered
electrical services for many of the roadway luminaires. With the use of LED
luminaires, it is important to start phasing out the unmetered electrical services so that
the actual energy savings can be harvested.
CONCLUSIONS
Over the lifetime of the luminaires, the use of LED luminaires will result in significant
savings in energy consumption and total lighting-related costs. The economic analysis
results suggested that the investment would be returned (compared to the base scenario) in 7
to 8 years for all scenarios, although scenario S2 (the existing HPS systems are converted
over 5 years) corresponded with the shortest investment return time. Over the 25-year
analysis period, the ROI for the different scenarios ranges between 3.25 and 5.76, with 3.25
projected for scenario 1; 4.45 for the 5-year replacement scenario; and 5.76 for the 10-year
65
replacement scenario. In addition, over the 25-year analysis period, converting HPS
luminaires to LED would cut the lighting energy cost at VDOT by 48% to 50% depending on
the conversion scenario. Due to the minimal maintenance required by LED luminaires,
replacing the traditional HPS luminaires will also significantly reduce the maintenance and
related costs.
LED luminaires outperformed the HPS system in light quality, distribution, and stability.
Overall, the LED systems exhibited much whiter light output, better light uniformity, and
lower glare and backlight. All LED luminaires evaluated emitted light that was much closer
in color natural light. In contrast, the HPS luminaires emitted yellowish light with high
special power within the 490 – 510 nm (green) and 560 – 620 nm (yellow) ranges. The LED
luminaires in general had a lower level of average horizontal and vertical illuminance
compared to the HPS luminaires. Note that this difference does not necessarily indicate
insufficient light output of the LED luminaires since lighting designs typically consider
multiple factors such as uniformity, light distribution, and BUG (backlight, uplight, and
glare) as required by different applications. The light output of most LED systems was much
more uniformly distributed over a larger area compared to that of the HPS system. Most
evaluated LED luminaires also exhibited a much more widely-spread vertical illuminance
distribution, indicating better vertical illuminance uniformity and less glare for travelers. In
addition, all LED luminaires evaluated showed a lower level of rear trespass light than the
HPS systems.
Compared to the LED systems, the HPS luminaires were much more prone to dirt
accumulation, particularly inside the lens due to poor ingress protection. Results also
suggested that the LED luminaires in general exhibited more stable performance (i.e., less
significant performance variations over time) than the HPS systems. The measured efficacy
of most LED luminaires was comparable to that of the HPS system, with Design C
luminaires slightly exceeding the HPS systems in efficacy.
Different LED designs showed differences in light distribution and lighting performance over
time. Among the different LED designs, luminaires with LED optic arrays (e.g., Design A,
B, and D) had better uniformity than luminaires with large LED optics (e.g., Design C). The
Design C luminaires with large LED light sources resembled the horizontal and vertical
illuminance distribution of the HPS systems more than the other LED designs. Overall, the
different LED designs exhibited very different horizontal illuminance distribution patterns,
with Design C and E more elongated while Design A, B, and D more circular. This suggests
that LED luminaires should be carefully chosen based on the application to result in the most
cost-effective lighting designs.
Exposed LED optic arrays did not attract more dirt than those covered with larger lenses
within the first 2 years of operation. However, it was more difficult to clean the individual
optics forming an array. The research team observed changes in horizontal light distribution
over time for Design A and B, likely due to irregular changes of the performance of
individual LED packages. The measured wattages of most LED luminaires were consistent
with manufacturer ratings with the single exception of Design D LED luminaires. The
66
Design B luminaire with the highest manufacturer-rated CCT was the only one that had a
CCT inconsistent with the manufacturer specification.
The light output of LED systems decreased during the first 2 years but not significantly when
performance variation and ambient factors are considered. Laboratory testing results
showed that the LED luminaires had 6% less output after 2 years of operation due to light
loss and dirt depreciation. The laboratory testing results suggested a much more significant
light loss due to lumen depreciation than due to dirt depreciation for LED luminaires.
Results also showed that dirt depreciation had a more significant impact on vertical
illuminance than on horizontal illuminance. Field data did not suggest a clear decrease in
horizontal illuminance over the 2 years, but the vertical illuminance decreased between 4%
and 12% for the LED luminaires. The lowest decrease in vertical illuminance was that of the
Design B luminaire, which had a three-panel folding design with the central panel facing
directly downwards and the other two folding towards the center.
The light color of most LED luminaires degraded over time during the 2 years of field
operation. This finding indicated that the color of the evaluated LED lights becomes
yellowish/reddish over time. The 2-year decrease in measured CCTs for all LED luminaires
was found to be between 2% and 8%.
LED fixtures need to be installed properly and have proper ingress protection. Data analysis
and visual inspection results suggested that, without proper installation procedures and/or
strict measurements, luminaires could fail to meet leveling and orientation requirements
during installation. In addition, installed luminaires may rotate and tilt over time if they are
not securely mounted. Many evaluated luminaires did not have housings with sufficient
ingress protection, resulting in significant rust/dirt accumulation and wildlife intrusion inside
the electrical compartment. The researchers did not find issues with the optical assembly
ingress protection for any of the LED designs. In contrast, the HPS system had significant
dirt and wildlife accumulation inside the optical lens.
RECOMMENDATIONS
1. VDOT’s TED should adopt LED technology for roadway lighting but implementation should
be project specific. LED roadway lighting must be project specific, and the project engineer
must review it on a project-by-project basis.
2. VDOT’s TED should develop and maintain a lighting inventory in the current/future asset
management framework. With the more widespread use of LED luminaires, it becomes
urgent to develop and implement a statewide lighting inventory system to manage, store, and
track the LED roadway luminaires and related files. Traditional luminaires (e.g., HPS) do
not involve warranty issues. In addition, retrofitting HPS luminaires is performed frequently
and at minimal costs. LED luminaires are relatively expensive, with a long service life (e.g.,
10 to 25 years); they vary in photometric performance (e.g., each product has a different light
distribution type, CCT, and/or fixture lumens); change rapidly (e.g., old models become
67
unavailable in a few years); and are warranted for a significant period (e.g., 100,000 hours).
Without an accurate and up-to-date lighting inventory system, VDOT will have difficulties
tracking the service lives of the luminaires against their warranties. In addition, VDOT will
likely lose original lighting design and technical files, resulting in difficulties in selecting
replacement luminaires without going through a new lighting design process. A good
lighting inventory system will also provide institutional information for energy consumption
and cost-benefit analyses if needed.
3. VDOT’s TED should update the VDOT Special Provision for LED Roadway Luminaires as
needed to reflect the latest technology status. Currently, roadway LED lighting technology is
rapidly improving. In the near future, many of the LED products on the market today will no
longer be available. Various new products with a potentially significantly different design
and materials will emerge, rendering many established performance metrics obsolete. As an
example, most LED luminaires acquired in 2012 for evaluation in this study were not in
production in late 2014 according to the venders contacted by the research team. Within this
context, VDOT should update the VDOT LED luminaire specifications as needed, with a
major revision every 4 to 5 years, until LED technology reaches its full maturity.
BENEFITS AND IMPLEMENTATION
Benefits
This study provides the needed know-how relevant to LED technology, the laboratory
and field performance of common LED roadway luminaires over time, and the expected cost-
benefit ratios if LED technology were implemented. The use of LED luminaires will result in
significant savings in energy consumption and total lighting-related costs. Over the 25-year
analysis period, the ROI for the different scenarios ranges between 3.25 and 5.76. Lighting
energy cost at VDOT after converting to LED lighting technology would be reduced by 48% to
50%.
Implementation
Implementation of the recommendations in the report is underway. VDOT is currently in
the process of phasing out the existing HPS luminaires on VDOT maintained roadways and
replacing them with LED luminaires. Based on the findings of this study, a VDOT Special
Provision for LED Roadway Luminaires document was developed to facilitate the selection of
LED luminaires for VDOT lighting projects. The following are the plan for and the status of
implementing the recommendations of this study:
Implementing LED technology and the VDOT Special Provision for LED Roadway
Luminaires. VDOT is currently in the process of implementing LED roadway
lighting technology statewide. Based on the findings of this study, the special
provision was finalized on July 23, 2015. VDOT Location and Design Division
(L&D) designers and consultants have begun using the special provision on design
68
projects. In addition, the special provision will be used for all future design projects
by October 2015. Statewide distribution of the special provision to VDOT staff for
use of the special provision was made on August 20, 2015.
Implementing a lighting inventory in the current/future asset management framework.
Currently, VDOT is expecting that a lighting inventory will be implemented as part of
VDOT’s Highway Maintenance Management System. The procurement process for
this system is underway.
Updating the VDOT Special Provision for LED Roadway Luminaires. VDOT’s TED
will determine the need to update the special provision at a future date as it deems
necessary. It is expected that the need will be reviewed no later than 4 to 5 years after
the special provision was implemented.
ACKNOWLEDGMENTS
The research team acknowledges the contribution of the following individuals during the
course of this research study: Mary Bennett, VDOT; Harry Campbell, VDOT; Benjamin Cottrell,
Jr., VDOT; Adam Dixon, VDOT; William Duke, VDOT; James Gillespie, VDOT; Edward
Hoppe, VDOT; Ning Li, VDOT; Marc Lipschultz, VDOT; Mansour Mahban, VDOT; Catherine
C. McGhee, VDOT; Audrey K. Moruza, VDOT; Vanloan Nguyen, VDOT; “Jon” Saeed Sayyar,
VDOT; Christopher Leone, Parsons Brinckerhoff; Paul Lutkevich, Parsons Brinckerhoff; Don
McLean, DMD & Associates Ltd.; and Kim Molloy, Parsons Brinckerhoff.
This research study was championed by TED and funded through the Virginia
Transportation Research Council, Virginia Department of Transportation.
REFERENCES
1. Building Technologies Office. Solid State Lighting: LEDs for Street Lighting – Here Today.
Document DOT/EE-1059. U.S. Department of Energy, Washington, D.C., November 2013.
2. Illuminating Engineering Society. Design Guide for Roadway Lighting Maintenance. IES
DG-4-14. New York, February 2014.
3. Office of Energy Efficiency and Renewable Energy. SSL Basics. U.S. Department of
Energy. http://energy.gov/eere/ssl/ssl-basics. Accessed December 1, 2014.
4. Ashe, M., Chwastyk, D., de Monasterio, C., Gupta, M., and Pegors, M. 2010 U.S. Lighting
Market Characterization. Nabigant Consulting, Inc., Chicago, January 2012.
5. Office of Efficiency & Renewable Energy. GATEWAY Demonstrations. U.S. Department of
Energy. http://energy.gov/eere/ssl/gateway-demonstrations. Accessed December 1, 2014.
69
6. Office of Efficiency & Renewable Energy. GATEWAY Demonstration Outdoor Projects.
U.S. Department of Energy. http://energy.gov/eere/ssl/gateway-demonstration-outdoor-
projects. Accessed December 1, 2014.
7. City of Sunnyvale. LED Street Lighting: City of Sunnyvale. Sunnyvale, Calif., 2009.
8. Li, S., Jiang, Y., Guan, B., Zhao, G., and Thompson, A. Cost- and Energy-Efficient (LED,
Induction and Plasma) Roadway Lighting. FHWA/IN/JTRP-2013/19. Joint Transportation
Research Program, Indiana Department of Transportation and Purdue University, West
Lafayette, 2013.
9. Pulugurtha, S.S. NC Roadway Lighting Needs Assessment, Maintenance Prioritization Tool
and Performance Measures. FHWA/NC/2012-14. University of North Carolina at
Charlotte, May 2013.
10. Long, S., Qin, R., Elmore, C., Ryan, T., and Schmidt, S. LED Roadway Luminaires
Evaluation. Final Report CMR 12-011. Research, Development and Technology, Missouri
Department of Transportation, Jefferson City, December 2011.
11. Office of Energy Efficiency and Renewable Energy. Solid-State Lighting Research and
Development: Multi-Year Program Plan. U.S. Department of Energy, Washington, D.C.,
April 2012.
12. Avrenli, K.A., Benekohal, R., and Medina, J. LED Roadway Lighting Volume 2: Field
Evaluations and Software Comparisons. FHWA-ICT-12-013. University of Illinois at
Urbana-Champaign, October 2012.
13. Virginia Department of Transportation, Operations Planning Office. Methods for
Calculating Maintenance and Operations Needs: Highway Lighting – Asset 380 (Needs
Methods Series: CY13-B.2). Richmond, June 2014.
14. Energy Management Solutions, Inc. Cost-Benefit Analysis of Energy Efficient Technologies
Available for Use in Roadway Lighting. Excelsior, Minn., December 2012.
15. U.S. Energy Information Administration. Short-Term Energy Outlook – Electricity.
February 10, 2015. http://www.eia.gov/forecasts/steo/report/electricity.cfm.
16. Harder, A. Energy Efficient Lighting. National Conference of State Legislatures, October
22, 2014. http://www.ncsl.org/research/energy/energy-efficient-lighting.aspx.
17. Dominion Virginia Power. Dominion Measures for Non-Residential Lighting Systems and
Control Program. https://www.dom.com/library/domcom/pdfs/ways-to-save-shared/non-res-
lighting-rebate-chart.pdf. Accessed February 11, 2015.
18. American Electric Lighting. High-Pressure Sodium Servicing Guide.
http://www.americanelectriclighting.com/Library/Literature/PDFs/HPS%20Servicing%20Gu
ide.pdf. Accessed April 15, 2015.
70
19. Ameren Illinois. Typical Lighting Wattages. ActOnEnergy Business Program.
http://www.actonenergy.com/portals/0/business/forms/lighting-wattage-guide.pdf. Accessed
April 15, 2015.
20. Pacific Northwest National Laboratory, SSL Pricing and Efficacy Trend Analysis for Utility
Program Planning. Pacific Northwest National Laboratory, Richland, Wash., October 2013.
21. Kimber, A., Roberts, J., Logan, J., and Lambert, M. LED Street Lighting: A Handbook for
Small Communities. Iowa Association of Municipal Utilities.
http://archive.iamu.org/services/electric/efficiency/Street%20Lighting/StreetLightingHandbo
ok.pdf. Accessed April 15, 2015.
22. American Electric Lighting. High Pressure Sodium Servicing Guide.
http://www.americanelectriclighting.com/Library/Literature/PDFs/HPS%20Servicing%20Gu
ide.pdf. Accessed February 18, 2015.
23. Illuminating Engineering Society. Roadway Lighting. ANSI/IES RP-8-14. New York,
2014.